Title of Invention	METHOD AND APPARATUS CONTROLLING POLYOLEFINE RESIN PRODUCTION PLANT
Abstract	The present invention relates to a control apparatus for a polyolefine resin production plant comprising: a production system capable of producing plural types of polyolefine resins; name data base means for previously registering names of a plurality of polyolefine resins producible by said production system as index data; pattern storage means for previously storing an operational pattern for an optimal type changing operation as an operational pattern table for each combination of a name of a current resin under production in a currently running production process by said production system and a name of a next target resin to be produced by said production system; pattern selecting means for comparing said name of said current resin with said name of said next target resin to discriminate a combination thereof and selecting an operational pattern corresponding to a discrimination result from said operational pattern table; and change control means for executing process control according to said operational pattern selected by said pattern selecting means to change a type of a resin to be produced by said production system.

Title of Invention

METHOD AND APPARATUS CONTROLLING POLYOLEFINE RESIN PRODUCTION PLANT

Abstract

The present invention relates to a control apparatus for a polyolefine resin production plant comprising: a production system capable of producing plural types of polyolefine resins; name data base means for previously registering names of a plurality of polyolefine resins producible by said production system as index data; pattern storage means for previously storing an operational pattern for an optimal type changing operation as an operational pattern table for each combination of a name of a current resin under production in a currently running production process by said production system and a name of a next target resin to be produced by said production system; pattern selecting means for comparing said name of said current resin with said name of said next target resin to discriminate a combination thereof and selecting an operational pattern corresponding to a discrimination result from said operational pattern table; and change control means for executing process control according to said operational pattern selected by said pattern selecting means to change a type of a resin to be produced by said production system.

Full Text	The present invention relates to a control apparatus for a Polyolefine resin production plant. Description of the Related Art In a production plant for producing polyolefines such as polyethylene and polypropylene, a distributed control system (hereinafter called "DCS") is typically used as a process control system. There has been a proposal to accomplish high-precision process control by coupling a process computer to this DCS via the system bus of the DCS. There are multifarious needs for polyolefine resins like polypropylene. In the production process in a production plant for polyolefine resins such as polypropylene, it is common to frequently change the names of resins (the types of resins) to be produced from one to another, e.g., from a high impact-strength resin having a high molecular weight to a high flow resin having a low molecular weight, or frequently change one polyolefine resin to a copolymer with another olefine. To meet such a demand, the present inventor proposed in Unexamined Japanese Patent Publication No. Sho 62-250010 the technique that detects a parameter concerning the physical property of polyethylene by means of a sensor or gas chromatography, performs operations on the detection value using a computer and controls the production of polyethylene having a predetermined physioal property based on the operational results. It is however difficult to adapt this technigue to the operaricnal control needed at the time one of multifarious types of polyolefine resins is changed to another. Parameters involved in changing resins from one. type to another in such a polyolefine resin production plant vary depending on the operational conditions. It is generally known to control the melt flow rate hereinafter called "MFR") by the amount of hydrogen supply to a polymerization reactor. To change one type of polyolefine resin under production to another in a short period of time for the purpose of increasing the hydrogen concentration in the reactor to reduce the polecular weight, it is efficient to supply a large amount of hydrogen at a time. This approach however ncreases the activity of the polymerization catalyst in accordance with the increase in hydrogen concentration. causing abnormal polymerization (e.g., an increase in polymerization temperature or polymerization reaction failure) originating from the rapid increase in hydrogen concentration. This is dangerous as well as undesirable because of the amount of polyolefine resins which do not meet the standards increases. Further, since the feed rate of hydrogen is closely associated with other parameters like the temperature of the reactor to be controlled, the conventional control by the DCS alone is very likely to produce a lots of products which do not meet the standards, i.e., defects, at the time of changing one type of resin to another. The suppression of the production of products which do not meet the standards therefor-e demands the knowledge-based or experience-based control by skillful operators. In changing one type of resin to another in a polyolefine resin production plant, conventionally, skillful operators changed a plurality of changeable elements based en their knowledge or experiences. Those changeable elements are common elements in process data, which should be changed according to a variation in production conditions and which, when changed, would cause other lots of process data to change. The common elements include the hydrogen concentration in a polymerization reactor-and a heat transfer coefficient. If such common elements in process data are found and are used as control parameters to control the operation of changing the types of resins from one to another, it should be possible to suppress the occurrence of products which do not meet the standards and to change the resin type properly and effectively. SUMMARY OF THE INVENTION Accordingly, it is an object of the present invention to provide a control method and apparatus for a polyolefine resin production plant, which, when needed, controls an ooeratioii of changing resins from one type to another in a polyolefine resin production plant using a proper process data element as a control parameter originatable from a change in production conditions, thereby suppressing the occurrence of products which do not meet the standards and ensuring the proper and effective automation for changing the resin type. A control method for a polyolefine resin production plant according to the first aspect of this invention comprises: a name registering step of previously registering names of a plurality of polyolefine resins producible by a production system into a name data base as index data; a pattern storing step of previously storing an operational pattern for an optimal type changing operation as an operational pattern stable into a pattern storage for each combination of a name of a current resin under production in a currently running production process by the production system and a name of a next target resin to be produced by the production system; a pattern selecting step of comparing the name of the current resin with the name of the next target resin to discriminate a combination thereof and selecting an operational pattern corresponding to a discrimination result from the operational pattern table; and a change control step of executing process control according to the operational pattern selected in the pattern selecting step to change a type of a resin to be produced by the production system. Each operation pattern to be stored in the operational pattern table may include a control target value for each resin type and information on a successive change in a feed rate of a raw material. The pattern selecting step may include a step of setting a start time for control by the selected operational pattern for a type changing operation to an optimal time for each operational pattern based on a given production end time for a type of a current resin. The production system may be used to produce multiple types of polyolefine resins different from one another in at least one of a molecular weight and a copolymer composition of polyolefine; and the pattern storing step may include a step of registering an optimal operational pattern given for each of a plurality of zones, obtained by previously dividing a hydrogen concentration in a polymerization reactor based on at least a relationship between a hydrogen concentration and activity of a polymerization catalyst, into the operational pattern table. In the previous case, the production system may be used to produce multiple types of polypropylene resins different from one another in at least one of a molecular weight and a copolymer composition of polyprony1ene. According to the second aspect of this invention, there is provided a control method for a polyolefine resin production plant for continuously changing a composition of a isupplied element in a polymerization reactor for continuous polymerization to thereby produce different types of polymers, which method comprises: a change control step of changing a control parameter based on an operational pattern by a changing operation model for each of previously stored resin names at a time one type of resin is changed to another in the continuous polymerization; a monitor step of always detecting and monitoring a sub-parameter which changes in accordance with a change in the control parameter; and a fuzzy control step of performing fuzzy control of an-operation based on the change control step in such^a manner that when a predetermined period of time passes after a change in the control parameter, the control parameter and the sub-parameter approach target values corresponding to a target resin type to be changed. The fuzzy control step may include: an operation step of performing a fuzzy operation )n a sub-parameter detected by the monitor step to )btain information on the control parameter; and a control step of changing the control parameter ibtained in the operation step. The predetermined period of time may be a time from point at which changing a resin type starts to a point t which the target values are set to values according o a target resin type to be changed after overaction of he control parameter. In the previous case, the method may further nclude a step of changing the control parameter to ontrol only a pressure in the polymerization reactor as he sub-parameter during the predetermined period of ime. The operational patterns may be obtained based on ast production achievements. The polyolefines may be product selected from olypropylenes, polyethylenes and polyethylene waxes. The control parameter may be a feed rate of ydrogen. The control parameter may be a feed rate of a atalyst. The control parameter may be a feed rate of at least one of a-olefines having 2-8 carbon atoms. The sub-parameter may be one of a ratio of hydrogen to propylene and a ratio of hydrogen to ethylene. The sub-parameter may be at least one of a pressure and a temperature in the polymerization reactor. A control apparatus for a polyolefine resin production plant according to the third aspect of this invention comprises: a production system capable of producing plural types of polyolefine resins; a name data base section fox~ previously registering names of a plurality of polyolefine resins producible by the production system as index data; a pattern storage foz previously storing an operational pattern for an optimal type changing operation as an operational pattern table for each combination of a name of a current resin under production in a currently running production process by the production system and a name of a next target resin to be produced by the production system; a pattern selector for comparing the name of the current resin with the name of the next target resin to discriminate a combination thereof and selecting an operational pattern corresponding to a discrimination result from the_^operational pattern table; and a change controller for executing process control according to the operational pattern selected by the pattern selector to change a type of a resin to be produced by the production system. The pattern storage may store an operation pattern including a control target value for each resin type and information on a successive change in a feed rate of a raw material in the operational pattern table. The pattern selector may further include a function of setting a start time for control by the selected operational pattern for a type changing operation to an optimal time for each operational pattern based on a gjven production end time for a type of a current resin. The production system may have a polymerization reactor and may produce multiple types of polyolefine resins different from one another in at least one of a molecular weight and a copolymer composition of polyolefine; and the pattern storage may register an optimal operational pattern given for each of a plurality of zones, obtained by previously dividing a hydrogen concentration in the polymerization reactor based on at least a relationship between a hydrogen concentration and activity of a polymerization catalyst, into the operational pattern table. In the previous case, the production system may produce multiple types of polypropylene resins different from one another in at least one_pf a molecular weight and a copolymer composition of polypropylene. A control apparatus for a polyolefine resin production plant according to the fourth aspect of this invention comprises: a production system having a polymerization reactor, for continuously changing a composition of a supplied element in the polymerization reactor for continuous polymerization to thereby produce different types of polymers; a pattern storage for storing an operational pattern by a changing operation model for each resin name; a change controller for changing a control parameter based on the operational patterns stored in the pattern storage to thereby change one type of resin to another in the continuous polymerization; a monitor section for always detecting and monitoring a sub-parameter which changes in accordance with a change in the control parameter; and a fuzzv controller for performing fuzzy control of the change controller in such a manner that when a predetermined period of time passes after a change in the control parameter, the control parameter and the sub-parameter approach target values corresponding to a target resin type to be changed. The fuzzy controller may include: an operation section for performing a fuzzy operation on a sub-parameter detected by the monitor section to obtain information on the control parameter; and a controller of changing the control parameter obtained in the operation section. The fuzzy controller may set the predetermined period of time as a time from a point at which changing a resin type starts to a point at which the target values are set to values according to a target resin type to be changed after overaction of the control parameter. In the previous case, the fuzzy controller may change the control parameter to control only a pressure in the polymerization reactor as the sub-parameter during the predetermined period of time. As the operational patterns, the pattern storage may store operational patterns obtained based on past production achievements- The production system may produce one of polvpropylenes, polyethylenes and polyethylene waxes as the polyolefines. The production system may use a feed rate of hydrogen as the control parameter. The production system may use a feed rate of a catalyst as the control parameter. The production system may use a feed rate of at least one of a-olefines having 2-8 carbons-atoms as the control parameter. The production system may use one of a ratio of hydrogen to propylene and a ratio of hydrogen to ethylene as the sub-parameter. The production system may use at least on of a pressure and a temperature in the polymerization reactor as the sub-parameter. According to the control method and control apparatus for a polyolefine resin production plant which embody this invention, the names of a plurality of polyolefine resins producible by the production system are previously registered into the name data base as index data, an operational pattern for an optimal type changing operation is previously stored as an operational pattern table into the pattex'n storage for each combination of the name of a current resin undei. production in a currently running production process by the production system and the name of a next target resin to be produced by the production system, an operational pattern corresponding to the name of the current resin and the name of a given next target resin to be produced is selected from the operational pattern table, and process control accoxding to the selected operational pattern is executed to change the type of a resin to be produced by the production system. This control system can suppress the occurrence of products which do not meet the standards and can accomplish the proper and effective automation for changing one type of resin to another. Accordingly, the present invention provides a control apparatus for a polyolefine resin production plant comprising: a production system capable of producing plural types of polyolefine resins; name data base means for previously registering names of a plurality of polyolefine resins producible by said production system as index data; pattern storage means for previously storing an operational pattern for an optimal type changing operation as an operational pattern table for each combination of a name of a current resin under production in a currently running production process by said production system and a name of a next target resin to be produced by said production system; pat.tern selecting means for comparing said name of said current resin with said name of said next target resin to discriminate a combination thereof and selecting an operational pattern corresponding to a discrimination result frcm said operational pattern table; and change control means for executing process control according to said operational pattern selected by said pattern selecting means to change a type of a resin to be produced by said production system. BRIEF DESCRIPTION OF THE DRAWINGS Other objects and advantages of the invention will become apparent during the following discussion in conjunction with the accompanying drawings, in which: FIG. 1 is a block diagram showing the structure of a control system for a polyolefine resin production plant according to a first embodiment of this invention; FIG. 2 is a functional block diagram for explaining the functional structure of the system in FIG. 1; FIG. 3 is a diagram showing the relationship between the hydrogen concentration and the activity of a polymerization catalyst in the system in FIG. 1; FIGS. 4A through 4D are diagrams showing different examples of a hydrogen supply operational pattern in the system in FIG. 1; FIG. 5 is a diagram showing the relationships between time and the hydrogen concentration for a first experimental example and a first comparative example in the system in FIG. 1; FIG. 6 is a diagram showing the relationships between time and the hydrogen concentration for a second experimental example and a second comparative example in the system in FIG. 1; FIG. 7 is a block diagram showing the structure of a control system for a polyolefine resin production plant according to a second embodiment of this invention; FIG. 8 is a diagram for explaining the concept of a model predictive control in the system in FIG. 7; FIG. 9 is a functional block diagram for explaining the functional structure of a part of the system in FIG. 7; FIG. 10 is a functional block diagram for explaining the functional structure of another part of the system in FIG. 7; FIG. 11 is a flowchart for explaining the operation of the system in FIG. 7; FIG. 12 is a diagram for explaining a control operation for a polymerization transition in the system in FIG. 7; and FIG. 13 is a diagram for explaining the control operation of the system in FIG. 7. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Control systems for a polyolefine resin production plant according to preferred embodiments of this invention will now be described with reference to the accompanying drawings. First Embodiment A control system for a polyolefine resin production plant according to the first embodiment of this invention will now be described with reference to The control system in FIG. 1 comprises a bus 1, a process computer 2, an operators station j, multifunction controllers 4, uninterrupted automatic controllers (UACs) 5, a wiring section 6, raw supply valves 7 and a magnetic disk apparatus 8. In FIG. 1, the process computer 2 and the operators station 3 are coupled to the bus 1. Those process computer 2 and operators station 3 perform the general control of the system. Further coupled to the bus 1 is the multifunction controller 4s which are controlled by the operators station 3. Control data for the multifunction controllers 4 are transferred to the respective uninterrupted automatic controllers 5 via the multifunction controllers 4, GO that even when the multifunction controleers 4 are disabled, the uninterrupted automatic controllers 5 can perform desired processing. The raw supply valves 7 for supplying a raw material to the plant are coupled via the wiring section 6 to the associated multifunction controllers 4, which control the opening and closing of the associated raw supply valves 7. More specific functions of the control system which is realized by the above hardware structure is exemplarily illustrated in FIG. 2, A name dat^ base 9 and an operational pattern table 10 shown in FIG. 2 are stored in storage areas set in the magnetic disk apparatus 8 as an external memory unit of the process computer 2 shown in FIG. 1. The names of about 100 types of resins, for example, are registered as indexes or index data in the name data base 9. The name data base 9 is a kind of an operation condition data base of each product type. For each type of resin, the specified values of plant operation process data such as the flow rate, temperature and pressure, and the target value for the product like MFR are registered in, for example, a sequential form in the name data base 9. Operational patterns each consisting of the operation values of a raw-material supply pattern are registered in the operational pattern table 10 each in association with the combination of the target value for a resin type and the target value for the next target resin to be changed. Each operational pattern is equivalent to the combination pattern of a composition value and a raw-material supply pattern. The process computer 2 compares current operation process data or the operation process data of the type of the current resin with the specified value of the operation process data of the type of the next resin, reads the optimal operational pattern from the operational pattern table 10 based on the combination of the compared operation process data, and controls the multifunction controllers 4 based on the operational pattern. When the end time for the current rype or resin has been set previously or the like, it is desirable that the process computer 2 should select the optimal operational pattern from the operational pattern table 10 and should select the start time for the optimal operational pattern for each selected operational pattern based on the end time for the current type of resin. The raw-material supply pattern used in this system will now be explained with the pattern of the feed rate hydrogen taken as an example. FIG. 3 shows the relationship between the hydrogen concentration (%) in a polymerization reactor and the activity of a polymerization catalyst (relative value) or a change in the activity of a polymerization catalyst with respect to a change in the hydrogen concentration in the polymerization reactor. The trend shown in FIG. 3 is true of the solution polymerization, slurry polymerization, bulk polymerization and gas-phase polrmerization. In FIG. 3, to reduce the molecular weight at the time of changing one type of resin to another, for example, it should be efficient to feed the sufficient rate of hydrogen at a time until a polymer with a low molecular weight is produced. As is apparent from FIG. 3, however, as the hydrogen concentration increases, the activity of the polymerization catalyst increases and such a rapid rise in hydrogen concentration causes abnormal polymerization. It is theriefore necessary to compare the hydrogen concentration at the operation value of the current type of resin with the target hydrogen concentration for the next type of resin to select the optimal supply pattern. According to this system, the hydrogen concentration in the reactor is previously divided into a plurality of zones based on the relationship between the hydrogen concentration and the activity of a polymerization catalyst, and the optimal operational pattern obtained for each segmented zone is registered in the operational pattern table 10. FIGS, 4A through 4D show the patterns of a change in the supply of a raw material or the patterns of a change in the feed rate of hydrogen for the individual segmented zones, which are to be registered in the operational pattern table 10. FIG. 4B shows the pattern which increases the feed rate of hydrogen in ancordance with the condition for the next type of resin and is suitable in the case where the difference between the hydrogen concentration in the operation value for the current resin type and that in the target value for the next resin type is less than 0.5%. The pattern shown in PrtS. 4C increases the feed rate of hydrogen in two steps and is suitable in the case where the hydrogen concentration in the operation value for the current resin type is less than 2% and the difference between the hydrogen concentration in the operation value for the current resin type and that in the target value for the next resin type is less than 0.5%. FIG. 4A shows the pattern which temporarily overfeeds the feed rate of hydrogen and is suitable in the case where the hydrogen concentration in the operation value for the current resin type lies off the range illustrated in TABLE 1. FIG. 4D shows the pattern for the case of reducing the feed rate of hydrogen. TABLE 1 shows what patterns should be selected based on the combination of the operation value for the current resin type and the target value for the next resin type, the rightmost column showing the feed rate of hydrogen per time. First Experimental bxampxe The type changing operation for changing one type of resin to another in the polymerization reactor was executed in the case where the hydrogen concentration in the operation value for the current type of resin was less than 2% and the hydrogen concentration in the operation value for the next type of resin was 4% or above. The process computer 2 read the operation value for the current type of resin from the distributed multifunction controllers 4, read the name of the next type of resin to be changed from the name data base 9, and derived the difference between the target values of the hydrogen concentration foi" both resins. As a result, the pattern in FIG. 4C was selected based on the TABLE 1. FIG. 4C shows the two-step hydrogen supply pattern in which the feed rate of hydrogen is increased to 20 Nm^/Hr and the feed rate of the catalyst is reduced at the first step, causing the polymerization amount in the polymerization reactor and thus reducing the catalyst concentration, and then the feed rate of hydrogen is further increased to 55 Nm^/Hr. When the hydrogen concentration reaches the target value, the feed rate of hydifogen is reduced to the specified value of 3.5 Nm^/Hr, after which the feed rate of hydrogen is adjusted to provide the target hydrogen concentration. It took 2 hours and 10 minutes for the physical property value of the produced polymer to meet the specified value under this control. First Comparative Example The same control as done in the first experimental example was executed with an operator manually operating only the distributed multifunction controllers. At the time of changing one type of resin to another, the operator simultaneously changed the set values of other multiple control loops and monitored those control loops, and performed the condition changing operation while discriminating the set values from the global point of view. Because the set values should be changed in full consideration of safety, the values were set to the minimum levels from the viewpoint of safety. In this case, to prevent abnormal polymerization, hydrogen was supplied for two hours at the rate of 20 NmVHr and the feed rate of hydrogen was then increased to 40 Nm^/Hr while monitoring the rise in hydrogen concentration. After confirming that the hydrogen concentration rose to the specified level, the feed rate of hydrogen was dropped to 3.5 Nm^/Hr. It took 3 hours and 45 minutes under this control _for the proper change of the type'of a resin or-for the physical property value of the product to meet the specified value. FIG. 5 shows changes in hydrogen concentration (%) of the first experimental example and the first comparative example. It is apparent from FIG. 5 that this invention can shorten the time for the physical property value of the next type of resin to reach the specified value. Second Experimental Example A product with a lower molecular weight was changed to a product with a higher molecular weight in the same system structure as used for the first experimental example. To lower the hydrogen concentration, it is effective to temporarily stop supplying hydrogen and perform dehydrogenation. The "dehydrogenation" means to perform degassing from the polymerization reactor to discharge hydrogen together with the reactive monomer gas. This method can considerably shorten the time needed to reduce the hydrogen concentration. As this method drops the pressure in the reactor which may affect the reaction, however, a large amount of hydrogen should not be discharged. In this respect, this should be used as an auxiliary scheme. In this experimental example, the process computer 2 selected the pattern in FIG. 4D, temporarily stopped the supply of hydrogen and performed dehydrogenation, and set the feed rate of hydrogen back to the specified value of 4.5 Nm^/Hr when the hydrogen concentration became the specified level. In this case, the changing of resins from one type to another took 3 hours- Second Comparative Example The operator manually performed the same control as done in the second experimental example. Because the operator should change the set values of other multiple control loops and monitor those control loops at the time of changing one type of resin to another in this case too, the operator performed only the operation to temporarily stop the supply of hydrogen first, and then returned the feed rate of hydrogen to the specified value of 4.5 Nm^/Hr. It took 5.5 hours to complete the proper change of the type of a I'esin in this case. FIG. 6 shows successive changes in hydrogen concentration (%) of the second experimental example and the second comparative example. It is apparent from FIG. 6 that this invention can shorten the time for the physical property value of the next type of resin to reach the specified value. The system shown in FIG. 1 may be realized by connecting a process computer to the bus of a conventional distributed control system so that this process computer accomplish the type change control. In-executing the type change control, the operation process data of the current type of resin is compared with the specified value of the operation process data of the next type of resin and the difference between both values is obtained to select the proper operational pattern. Operational patterns to be registered in the operational pattern table are what is need for the operation process data of the plant, after changed, to reach the specified value of the operation process data of the next type of resin to be produced in order to ensure the continuous transition to the operation for the next type of resin without stopping the operation even after the completion of the operation for the current type of resin. They are considered the optimal operational patterns in that the quantity of products which do not meet the standards and are produced at the time of changing the types of resins and in that the transition to the operation of producing the next type of resin is accomplished safely and efficiently. The operational patterns are obtained by dividing the conditions for changing operation process data into several types of patterns. As the target value of each type of resin and a successive change in the feed rate of each raw material are registered in the operational pattern table, it is possible to register the operational pattern which is based on -the result af—the analysis previously done through the manual operation for each type of resin. This can ensure the control that does not depend on the skill levels of the individual operators. Second Embodiment A control system for a polyolefine resin production plant according to the second embodiment of this invention will be described below. The second embodiment is adapted to a continuous polymerization system for polymers which automatically changes polymers from one type to another in the continuous polymerization. Polymers or high-molecular materials are produced by polymerization. Polymers can be produced by continuous polymerization in the polymerization reactor, This continuous polymerization sequentially produces plural types of polymers which have different molecular weights and different densities. The switching of the composition in the polymerization reactor should be executed promptly and stably in the continuous polymerization in order to reduce the number of transition products (the off-standards products which are produced transiently e the time of switching from one type of polymer to another) as much as possible when one polymer is changt to another. For example, the ratio of feed rates of supplied gases, such as the raJ;io of the feed rates of hydrogen ethylene and propylene in changing the molecular weigh or the ratio of the feed rates of propylene and ethylene in changing the molecular weight/density corresponds to a change in composition in the polymerization reactor and becomes an index for changing the polymer type or the transition. There is one conventional continuous polymerization technique which over-supplies a catalyst, hydrogen or propylene for a certain period of time in order to cause the ratio of the value measured by a gas chromatography (hereinafter called "gas chromatographic ratio") to approach the target value for the type of a transition product by adjusting the feed rate of the catalyst, the feed rate of hydrogen, the feed rate of propylene or the like with the feed rate of ethylene kept constant, and then returns that ratio to said target value. As already mentioned, conventionally, switching the polymer type is manually performed by an operator and thus greatly depends on the knowledge- and experience-based skill of that operator. Accordingly, the time for the transition varies depending on operators, thus making it difficult to stabilize the qualities of products around the transition for some operators. In this respect, the second embodiment of this invention provides a continuous polymerization system for polymers which is capable of automatically and quickly switches one type of polymer to another in the continuous polymerization while improving the qualities of the products. That is, the second embodiment is a continuous polymerization system for polymers to which the control system of this invention for a polyolefine resin production plant is applied. FIG. 7 shows the hardware structure of the second embodiment. The continuous polymerization system for polymers sequentially produces different types of polymers through the continuous polymerization which is accomplished by continuously switching compositions in the polymerization reactor. The continuous polymerization system for polymers in FIG. 7 comprises a model algorithm control section (hereinafter called "MAC section") 21, a distributed control system section (hereinafter called "DCS section") 23, and a polymerization reactor 28. The MAC section 21 sets a model pattern for the DCS section 23 and performs fuzzy control. Accordingly, the MAC section 21 manages model pattern parameters fuzzy parameters and a name condition file and manages the fuzzy control. The DCS section 23 executes model-pattern based program control to perform the general control of the system and the sequence control. The MAC section 21 executes the model predictive control using a model pattern- obtained by patterning the overaction part in a test sequence prepared by the DCS section 23. The "model pattern" is the operational pattern of the feed rates of hydrogen and propylene set on the basis of the past achievements for the individual polymer types. A scheme of patterning the transitional operations based on the experiences of skillful operators to construct models for the individual polymer types belongs to the model predictive control. For the optimization purpose, a control apparatus which employs the model predictive control should be used. The model predictive control includes MAC (Model Algorithmic Control) and DMC (Dynamic Matrix Control), for example. The typical method for the model predictive control uses the dynamic characteristic model of the process in the control algorithm to predict the future behaviors Yp(t), Yp(t+1), ... of a control target from the previous operational data and to control it within the finite time, as illustrated in FIG. 8. With the use of the model pattern alone, the fuzzy control in the MAC section 21 cannot cope with the control of the physical property which changes from time to time. Therefore, the fuzzy control is used to control the physical property to the target value. In the fuzzy control, membership functions and fuzzy rules are set in association with the manual control operation so as to achieve a human-like control operation. The continuous polymerization system for polymers performs a fuzzy operation on the process feedback values and reflects the operation results on the process again. More specifically, the continuous polymerization system for polymers executes sequence control which performs control based on model patterns attained from the past operational achievement data for the individual polymer types and indirect advanced control which performs a fuzzy operation at a higher level than the sequence control and sets the operation results in the DCS section 23 for advanced control. The DCS section 23 may be constructed by using a general-purpose computer system (e.g., "CENTUM" (product of YOKOGAWA ELECTRIC CO.) or "TDCS" (product of YAMATAKE HONEYWELL CO.)),, which includes an operators station 24, an operators console 25 and a field controller 26. The field controller 26 has a batch setter 55 to be discussed later. The MAC section 21 has a hard disk drive, a random access memory and an RS71 interface card, though not illustrated, and operates the system based on software. The software includes a line controller, BASIC interpreter, an interface card utility, an interactive graphics utility, a trend library, a data base utility, DOS (Disk Operating System), a conversion utility, a shell and a fuzzy processing program (such as "AdMAS" which is a packaged program). A catalyst, hydrogen, ethylene and propylene Dnomer or the like are properly supplied to the Dlymerization reactor 28 for polymerization. The field ontroller 26 receives data such as the feed rate of the atalyst, the feed rate of hydrogen, the feed rate of ropylene and the like. The field controller 26 ontrols the pressure and the temperature in the olymerization reactor 28, and the gas chromatographic atio. The polymerization reactor 28 performs continuous lolymerization. The polymer may be polyolefine which is i high-molecule compound obtained by the polymerization )f a-olefines each of which has a carbon-carbon double )ond in the molecules. Polyolefine includes low-density )olyethylene which is synthesized by radical polymerization. The polyolefine may also be polyethylene having an intermediate or high density or a polyolefine wax. The functional structure of the continuous polymerization system for polymers that can be achieved by software will now be described. FIG. 9 shows the functional structure of the MAC section 21, and FIG. 10 shows the functional structure of the DCS section 23. The software of the continuous polymerization system for polymers consists of the MAC function portior shown-in FIG. 9 and the DCS sequence function portion also shown in FIG. 9. The MAC function portion include! a data base 29, a fuzzy control section 31, a trend monitor section 32, a data base input/output section 33, a data collecting section 36, a data setting section 37, a fuzzy processor 38, a simple simulator 41, a transition data collecting section 42, a transition setting section 43, a file selector 44 and a data analysis/trend section 45. The fuzzy control section 31 manages the data input/output of the fuzzy processor 38 and name condition parameters, and performs fuzzy control like data processing and operation and pressure control of model patterns. In other words, the fuzzy control section 31 executes fuzzy control in such a way that sub-parameters like the gas chi-omatographic ratio, pressure and temperature, which change in accordance with the values of control parameters like the feed rate of hydrogen and the feed rate of a catalyst, are permitted to approach the target values associated with the next type of polymer to be changed, upon passage of a predetermined time passes after the values of the control parameters change. The trend monitor section 32 presents the trend display of the transition status in real time to displa; the contents and status of the fuzzy control- The data base input/output section 33 manages the input/output access to a data file 29A in the data base 29, such as the aforementioned model pattern parameters and name operation condition data, and inputs and changes those data. The simple simulator 41 has a process simulation function to store a simple model equation and a processing model equation of the process simulation, infer the physical property of a polymer and inspect the fuzzy output. The data collecting section 36 collects data from the DCS section 23 through the communications software that is invoked on the line controller and stores the data in a common area. That is, the data collecting section 36 serves as a detecting section to detect the values of the aforementioned sub-parameters. The fuzzy processor 38 incorporates a fuzzy operation section which is provided as a subprogram of the fuzzy processing program, and executes a fuzzy operation under the control of the fuzzy control section 31. That is, the values o£ the aforementioned control parameters are attained from the fuzzy operation on the values of the sub-parameters that are collected by the data collecting section 36. The data setting section 37 sets the model pattern data of the type of a resin and the values obtained through the fuzzy operation into the batch setter 55 of the field controller 26. The fuzzy processing progr.ajn builds fuzzy rules, membership functions and the like. Each fuzzy rule has a condition (IF) part and a conclusion (THEN) part each consisting of a fuzzy set. The following are some examples of fuzzy rules. Rule 1 IF X is Aj and y is Bj THEN z is Cj Rule 2 IF X is A2 and y is B2 THEN z is Cj Both rules have two condition part variables "x is A" and "y is B" and one conclusion part variable "z is C." In the rules, Aj^, Aj, B^, B2, C^ and C2 are fuzzy sets. The membership functions define the parts of a fuzzy rule for quantization. The fuzzy operation obtains the matchability (goodness of fit) of the condition part of each rule with given values, obtains the inference result of each rule based on the attained matchability, and then attains the final inference result from the inference result of each rule. The simple simulator 41 inputs data from the transition data file 29A to the fuzzy processor 38, provides the trend display of the output values on the trend monitor section 32 and adjusts the setting of the membership functions and fuzzy rules for the fuzzy control based on the displayed results. The transition data collecting section 42 stores, for example, 88 pieces of data from one hour before th beginning of the transition up to the end of the transition into the data file 29A in the data base 29. The transition setting section 43 sets the trend name and the time for starting the data collection. The file selector 44 selects the transition data file 29A to be displayed. The data analysis/trend section 45 presents the trend display of the transition data of the designated name in the data file 29A. Then, the DCS sequence section 50 has a system control sequence section 51, a parameter setting sequence section 52, a fuzzy control sequence section 53, a model pattern sequence section 54 and the aforementioned batch setter 55. The fuzzy control sequence section 53 and model pattern sequence section 54 control the polymerization reactor 28. The system control sequence section 51 controls the selection of the resin type, the selection of the sequence in use, the start and stop of the sequence, the selection of a model pattern parameter and the communication with the MAC section 21. The resin type is managed by an 8-bit code; a binary code is handled with a type number in a decimal notatioh- The parameter setting sequence section 52 communicates with the MAC section 21 upon the initiation of the sequence after a resin type is selected, and sets model pattern parameters in the batch setter 55 of the DCS section 23. The parameter setting sequence section 52 sets the same value as two model pattern parameters in the batch setter 55. This sequence section 52 compares the two values, set in the batch setter 55, with each other and determines that the proper communication has been accomplished when those two values are the same, and sets this value in a program setter (not shown). When the two model pattern parameters in the batch setter 55 differ from each other, the sequence section 52 sends a communication error message to the MAC section 21 and retries the communication. The fuzzy control sequence section 53 manages the fuzzy operation values which are to be set by the data setting section 37 of the MAC section 21 and causes the batch setter 55 to write in a control tag- This sequence section 53 is enabled when a predetermined period of time passes after the setting of the model pattern parameters, and controls the communication between the MAC section 21 and the DCS section 23. The model pattern sequence section 54 is a patterning sequence which optimizes the patterning and constructs a model pattern for the resin types whose fuzzy parameters have not been built yet. This sequence section 54 changes the values of control parameters based on the moder^attern for each type at the time of switching one type of polymer to another in the continuous polymerization. The control operation for the continuous polymerization in the continuous polymerization system for polymers will now be discussed. FIG. 11 is a flowchart illustrating the continuous polymerization control, and FIG. 12 illustrates the control characteristic of the polymerization transition. First, the MAC section 21 builds model patterns (step SOO). In this case, the test pattern sequence performed patterning for 18 types, for example. Further, the patterning was performed with respect to 3 parameters: the feed rate of hydrogen, the feed rate of a catalyst and the feed rate of propylene. FIG. 13 exemplifies a model pattern for hydrogen. As shown in FIG. 13, the hydrogen pattern indicates the hydrogen supply of 30 Nm^/Hr for 25 minutes and the hydrogen supply of 120 Nm^/Hr for 30 minutes at the time of overaction. With regard to the catalyst, the activity always varies, making it difficult to previously determine the feed rate. Therefore, the ratio of a change in activity is factorized based on the past achievements, and the feed rate is computed by the MAC section 21 spontaneously upon the activity at the transition time. The computed feed rate is set in the DCS section 23. _ Because the constructed model pattern has been-optimized to a certain degree from the past aohievements, the polymerization transition time can be shortened. For instance, the time needed for the transition work (including feed-down and feed-up) is about 12 hours on average. The automation system for transition polymerization shortens this transition time to 10 hours, two hours shorter. The setting of the transition type, and the start and stop of the sequence are executed on the screen of the operators station 24 and a message is sent to the MAC section 21 from the DCS section 23 (step SOI), The MAC section 21 receives the message from the DCS section 23 and sends an acknowledge message to the DCS section 23 (step S02). Next, the MAC section 21 determines if the message from the DCS section 23 is corx-ect (step 503). VJhen the message from the DCS section 23 is correct, the data setting section 37 in the MAC section 21 sets model pattern parameters in the batch setter 55 in the DCS section 23 (step 304). In this example, two same model pattern parameters are set in the batch setter 55. Then, the DCS section 23 determines if the set model pattern parameters have been transferred by comparing a flag with the two same model pattern parameters (step S05), When the model pattern parameters have been transferred properly, it is determined if fuzzy setting of the type has been performed (step S06). When this fuzzy setting has been done, the system control sequence section 51 selects and invokes the fuzzy control sequence section 53 (step S07). When the fuzzy control sequence section 53 is selected, the transition operation is initiated with the previously defined model pattern and is switched to fuzzy control after a predetermined period of time (step SOS). More specifically, overaction is performed to change the feed condition to the condition for the next type of product in the initial state at the time of transition, and after the passage of the predetermined period of time, fuzzy control is performed to set the gas flow rate to the H/E target value shown in FIG. 12. The "predetermined period of time" is the time for setting the control target value to the value for the next type of product upon the execution of the overaction after the initiation of the transition, and is set for each type. In this case, the fuzzy control section 31 executes the gas chromatography control (H/E control) and pressure control based on the control algorithm during the fuzzy control period as shown in FIG. 12. The control algorithm can be changed for fine adjustment even during operation. The control cycle is separated to two paxts for the gas chromatography control and pressure control, each of which can be set arbitrarily. For example, the chromatography control is carried out every 3 minutes while the pressure control is executed every minute. When the physical property reaches the target valu( and becomes stable, the transition is ended and the normal operation is initiated (step S09). The fuzzy control continues even in the normal operation to perform stabilization control in this operation. The fuzzy control sequence continues until the feed-up, the feed-down and the next transition. When the fuzzy control sequence is stopped arbitrarily based on the control table, control is switched to the ordinary control by the manual operation. When fuzzy setting has not bee made, the system control sequence section 51 selects and invokes the model pattern sequence section 54 (step SIO). This sequence section 54 starts the model pattern control shown in FIG. 12 at the time of overaction (ste Sll). In the pattern compensation in this case, the pressure of n minutes later is inferred for the pressui control, Since overaction is taken in the predetermined period of time, the physical property is not controllec and only the pressure in the polymerization reactor is controlled. The data collecting section 36 of *he high-rank MJ section 21 collects, for example, five pieces of pressure data each obtained every minure ana ou-cputs xne ratio of change in the collected pressure data (step S12). Pressure change for one minute: dpl=P-Pl, dp2=Pl-P2, dp3=P2-P3, dp4=P3-P4, dp5=P5-P6 Then, the data collecting section 36 infers pressure of one minute later from the current point, based on the data on the ratio of change (step S13). Inferred pressure after one minute: AVPA=SUM(dpl, dp2, dp3, dp4, dp5)/5 +reactor pressure D201PS The data collecting section 36 determines if the inferred pressure value has reached the pressure point (upper limit/lower limit) previously determined by the type condition (step S14). At this point of time, the data collecting section 36 changes the amount of the catalyst to control the pressure (step S15). The amount of a single operation of the feed rate of the catalyst was the finite difference between the first point and the second point in the model pattern based on the past achievements. When the pressure reaches the point determined by the type condition, the ^4AC section 21 outputs an upper-limit pressure signal and a lower-limit pressure signal to the DCS sectio^n 23 to perform the sequence-based control (step S16). Since the gas chromatography control is not executed during this period, the manual operation is necessary when the operation enters the stable state. The pattern sequence is used to optimize the model pattern, not for automation. The automation of the polymerization transition i carried out by the construction of the model patterns and the fuzzy control. The pattern of the test sequen which has constructed the model patterns under the model-pattern based program control can be automatical set by the MAC section 21 to be used for the optimization of the type-by-type patterns. It is therefore possible to develop the scheme oi constructing the fuzzy parameters for the automatic switching of polymers as needed, even for other types polymers than the typical polymers which are automatically changed from one to another. During the predetermined period of time, only th The system may also use a feed rate of at least of a-olefines having 2-8 carbon atoms, e.g., jethylene propylene, butylene and the like, as the control parameter. It is apparent that, in this invention, a wide range of different working modes can be formed based on the invention without deviating from the spirit and scope of the invention. WE CLAIM; 1 . A control apparatus for a polyolefine resin production plant comprising: a production system capable of producing plural types of polyolefine resins; name data base means for previously registering names of a plurality of polyolefine resins producible by said production system as index data; pattern storage means for previously storing an operational pattern for an optimal type changing operation as an operational pattern table for each combination of a name of a current resin under production in a currently running production process by said production system and a name of a next target resin to be produced by said production system; pattern selecting means for comparing said name of said current resin with said name of said next target resin to discriminate a combination thereof and selecting an operational pattern corresponding to a discrimination result from said operational pattern table; and change control means for executing process control according to said operational pattern selected by said pattern selecting means to change a type of a resin to be produced by said production system. 2 . The control apparatus according to claim 1, wherein said pattern storage means has means for storing an operation pattern having a control target value for each resin type and information on a successive change in a feed rate of a raw material in said operational pattern table. 3. The control apparatus according to claim 2, wherein said pattern selecting means has means for setting a start time for control by said selected operational pattern for a type changing operation to an optimal time for each operational pattern based on a given production end time for a type of a current resin. 4. The control apparatus according to claim 2, wherein production system has a polymerization reactor to produce different types of polyolefine resins having different molecular weight and a copolymer composition of polyolefine; and said pattern storage means to register an optimal operational pattern given for each of a plurality of zones, obtained by previously dividing a hydrogen concentration in said polymerization reactor based on at least a relationship between a hydrogen concentration and activity of a polymerization catalyst, into said operational pattern table. 5. A control apparatus for a polyolefine resin production plant, substantially as herein described with reference to the accompanying drawings.

Full Text

The present invention relates to a control apparatus for a Polyolefine resin production plant.
Description of the Related Art
In a production plant for producing polyolefines such as polyethylene and polypropylene, a distributed control system (hereinafter called "DCS") is typically used as a process control system. There has been a proposal to accomplish high-precision process control by coupling a process computer to this DCS via the system bus of the DCS.
There are multifarious needs for polyolefine resins like polypropylene. In the production process in a production plant for polyolefine resins such as polypropylene, it is common to frequently change the names of resins (the types of resins) to be produced from one to another, e.g., from a high impact-strength resin having a high molecular weight to a high flow resin having a low molecular weight, or frequently

change one polyolefine resin to a copolymer with another olefine.
To meet such a demand, the present inventor proposed in Unexamined Japanese Patent Publication No. Sho 62-250010 the technique that detects a parameter concerning the physical property of polyethylene by means of a sensor or gas chromatography, performs operations on the detection value using a computer and controls the production of polyethylene having a predetermined physioal property based on the operational results. It is however difficult to adapt this technigue to the operaricnal control needed at the time one of multifarious types of polyolefine resins is changed to another.
Parameters involved in changing resins from one. type to another in such a polyolefine resin production plant vary depending on the operational conditions.
It is generally known to control the melt flow rate hereinafter called "MFR") by the amount of hydrogen supply to a polymerization reactor. To change one type of polyolefine resin under production to another in a short period of time for the purpose of increasing the hydrogen concentration in the reactor to reduce the polecular weight, it is efficient to supply a large amount of hydrogen at a time. This approach however ncreases the activity of the polymerization catalyst in accordance with the increase in hydrogen concentration.

causing abnormal polymerization (e.g., an increase in polymerization temperature or polymerization reaction failure) originating from the rapid increase in hydrogen concentration. This is dangerous as well as undesirable because of the amount of polyolefine resins which do not meet the standards increases.
Further, since the feed rate of hydrogen is closely associated with other parameters like the temperature of the reactor to be controlled, the conventional control by the DCS alone is very likely to produce a lots of products which do not meet the standards, i.e., defects, at the time of changing one type of resin to another. The suppression of the production of products which do not meet the standards therefor-e demands the knowledge-based or experience-based control by skillful operators.
In changing one type of resin to another in a polyolefine resin production plant, conventionally, skillful operators changed a plurality of changeable elements based en their knowledge or experiences. Those changeable elements are common elements in process data, which should be changed according to a variation in production conditions and which, when changed, would cause other lots of process data to change. The common elements include the hydrogen concentration in a polymerization reactor-and a heat transfer coefficient.
If such common elements in process data are found

and are used as control parameters to control the operation of changing the types of resins from one to another, it should be possible to suppress the occurrence of products which do not meet the standards and to change the resin type properly and effectively.
SUMMARY OF THE INVENTION
Accordingly, it is an object of the present invention to provide a control method and apparatus for a polyolefine resin production plant, which, when needed, controls an ooeratioii of changing resins from one type to another in a polyolefine resin production plant using a proper process data element as a control parameter originatable from a change in production conditions, thereby suppressing the occurrence of products which do not meet the standards and ensuring the proper and effective automation for changing the resin type.
A control method for a polyolefine resin production plant according to the first aspect of this invention comprises:
a name registering step of previously registering names of a plurality of polyolefine resins producible by a production system into a name data base as index data;
a pattern storing step of previously storing an operational pattern for an optimal type changing operation as an operational pattern stable into a pattern storage for each combination of a name of a current

resin under production in a currently running production process by the production system and a name of a next target resin to be produced by the production system;
a pattern selecting step of comparing the name of the current resin with the name of the next target resin to discriminate a combination thereof and selecting an operational pattern corresponding to a discrimination result from the operational pattern table; and
a change control step of executing process control according to the operational pattern selected in the pattern selecting step to change a type of a resin to be produced by the production system.
Each operation pattern to be stored in the operational pattern table may include a control target value for each resin type and information on a successive change in a feed rate of a raw material.
The pattern selecting step may include a step of setting a start time for control by the selected operational pattern for a type changing operation to an optimal time for each operational pattern based on a given production end time for a type of a current resin.
The production system may be used to produce multiple types of polyolefine resins different from one another in at least one of a molecular weight and a copolymer composition of polyolefine; and the pattern storing step may include a step of registering an optimal operational pattern given for each of a

plurality of zones, obtained by previously dividing a hydrogen concentration in a polymerization reactor based on at least a relationship between a hydrogen concentration and activity of a polymerization catalyst, into the operational pattern table.
In the previous case, the production system may be used to produce multiple types of polypropylene resins different from one another in at least one of a molecular weight and a copolymer composition of polyprony1ene.
According to the second aspect of this invention, there is provided a control method for a polyolefine resin production plant for continuously changing a composition of a isupplied element in a polymerization reactor for continuous polymerization to thereby produce different types of polymers, which method comprises:
a change control step of changing a control parameter based on an operational pattern by a changing operation model for each of previously stored resin names at a time one type of resin is changed to another in the continuous polymerization;
a monitor step of always detecting and monitoring a sub-parameter which changes in accordance with a change in the control parameter; and
a fuzzy control step of performing fuzzy control of an-operation based on the change control step in such^a manner that when a predetermined period of time passes

after a change in the control parameter, the control parameter and the sub-parameter approach target values corresponding to a target resin type to be changed.
The fuzzy control step may include:
an operation step of performing a fuzzy operation )n a sub-parameter detected by the monitor step to )btain information on the control parameter; and
a control step of changing the control parameter ibtained in the operation step.
The predetermined period of time may be a time from point at which changing a resin type starts to a point t which the target values are set to values according o a target resin type to be changed after overaction of he control parameter.
In the previous case, the method may further nclude a step of changing the control parameter to ontrol only a pressure in the polymerization reactor as he sub-parameter during the predetermined period of ime.
The operational patterns may be obtained based on ast production achievements.
The polyolefines may be product selected from olypropylenes, polyethylenes and polyethylene waxes.
The control parameter may be a feed rate of ydrogen.
The control parameter may be a feed rate of a atalyst.

The control parameter may be a feed rate of at least one of a-olefines having 2-8 carbon atoms.
The sub-parameter may be one of a ratio of hydrogen to propylene and a ratio of hydrogen to ethylene.
The sub-parameter may be at least one of a pressure and a temperature in the polymerization reactor.
A control apparatus for a polyolefine resin production plant according to the third aspect of this invention comprises:
a production system capable of producing plural types of polyolefine resins;
a name data base section fox~ previously registering names of a plurality of polyolefine resins producible by the production system as index data;
a pattern storage foz previously storing an operational pattern for an optimal type changing operation as an operational pattern table for each combination of a name of a current resin under production in a currently running production process by the production system and a name of a next target resin to be produced by the production system;
a pattern selector for comparing the name of the current resin with the name of the next target resin to discriminate a combination thereof and selecting an operational pattern corresponding to a discrimination result from the_^operational pattern table; and
a change controller for executing process control

according to the operational pattern selected by the pattern selector to change a type of a resin to be produced by the production system.
The pattern storage may store an operation pattern including a control target value for each resin type and information on a successive change in a feed rate of a raw material in the operational pattern table.
The pattern selector may further include a function of setting a start time for control by the selected operational pattern for a type changing operation to an optimal time for each operational pattern based on a gjven production end time for a type of a current resin.
The production system may have a polymerization reactor and may produce multiple types of polyolefine resins different from one another in at least one of a molecular weight and a copolymer composition of polyolefine; and the pattern storage may register an optimal operational pattern given for each of a plurality of zones, obtained by previously dividing a hydrogen concentration in the polymerization reactor based on at least a relationship between a hydrogen concentration and activity of a polymerization catalyst, into the operational pattern table.
In the previous case, the production system may produce multiple types of polypropylene resins different from one another in at least one_pf a molecular weight and a copolymer composition of polypropylene.

A control apparatus for a polyolefine resin production plant according to the fourth aspect of this invention comprises:
a production system having a polymerization reactor, for continuously changing a composition of a supplied element in the polymerization reactor for continuous polymerization to thereby produce different types of polymers;
a pattern storage for storing an operational pattern by a changing operation model for each resin name;
a change controller for changing a control parameter based on the operational patterns stored in the pattern storage to thereby change one type of resin to another in the continuous polymerization;
a monitor section for always detecting and monitoring a sub-parameter which changes in accordance with a change in the control parameter; and
a fuzzv controller for performing fuzzy control of the change controller in such a manner that when a predetermined period of time passes after a change in the control parameter, the control parameter and the sub-parameter approach target values corresponding to a target resin type to be changed.
The fuzzy controller may include:
an operation section for performing a fuzzy operation on a sub-parameter detected by the monitor

section to obtain information on the control parameter; and
a controller of changing the control parameter obtained in the operation section.
The fuzzy controller may set the predetermined period of time as a time from a point at which changing a resin type starts to a point at which the target values are set to values according to a target resin type to be changed after overaction of the control parameter.
In the previous case, the fuzzy controller may change the control parameter to control only a pressure in the polymerization reactor as the sub-parameter during the predetermined period of time.
As the operational patterns, the pattern storage may store operational patterns obtained based on past production achievements-
The production system may produce one of polvpropylenes, polyethylenes and polyethylene waxes as the polyolefines.
The production system may use a feed rate of hydrogen as the control parameter.
The production system may use a feed rate of a catalyst as the control parameter.
The production system may use a feed rate of at least one of a-olefines having 2-8 carbons-atoms as the control parameter.

The production system may use one of a ratio of hydrogen to propylene and a ratio of hydrogen to ethylene as the sub-parameter.
The production system may use at least on of a pressure and a temperature in the polymerization reactor as the sub-parameter.
According to the control method and control apparatus for a polyolefine resin production plant which embody this invention, the names of a plurality of polyolefine resins producible by the production system are previously registered into the name data base as index data, an operational pattern for an optimal type changing operation is previously stored as an operational pattern table into the pattex'n storage for each combination of the name of a current resin undei. production in a currently running production process by the production system and the name of a next target resin to be produced by the production system, an operational pattern corresponding to the name of the current resin and the name of a given next target resin to be produced is selected from the operational pattern table, and process control accoxding to the selected operational pattern is executed to change the type of a resin to be produced by the production system. This control system can suppress the occurrence of products which do not meet the standards and can accomplish the proper and effective automation for changing one type of

resin to another.
Accordingly, the present invention provides a control apparatus for a polyolefine resin production plant comprising:
a production system capable of producing plural types of polyolefine resins;
name data base means for previously registering names of a plurality of polyolefine resins producible by said production system as index data;
pattern storage means for previously storing an operational pattern for an optimal type changing operation as an operational pattern table for each combination of a name of a current resin under production in a currently running production process by said production system and a name of a next target resin to be produced by said production system;
pat.tern selecting means for comparing said name of said current resin with said name of said next target resin to discriminate a combination thereof and selecting an operational pattern corresponding to a discrimination result frcm said operational pattern table; and
change control means for executing process control according to said operational pattern selected by said pattern selecting means to change a type of a resin to be produced by said production system.

BRIEF DESCRIPTION OF THE DRAWINGS
Other objects and advantages of the invention will become apparent during the following discussion in conjunction with the accompanying drawings, in which:
FIG. 1 is a block diagram showing the structure of a control system for a polyolefine resin production plant according to a first embodiment of this invention;
FIG. 2 is a functional block diagram for explaining the functional structure of the system in FIG. 1;
FIG. 3 is a diagram showing the relationship between the hydrogen concentration and the activity of a polymerization catalyst in the system in FIG. 1;
FIGS. 4A through 4D are diagrams showing different examples of a hydrogen supply operational pattern in the system in FIG. 1;
FIG. 5 is a diagram showing the relationships between time and the hydrogen concentration for a first experimental example and a first comparative example in the system in FIG. 1;
FIG. 6 is a diagram showing the relationships between time and the hydrogen concentration for a second experimental example and a second comparative example in the system in FIG. 1;
FIG. 7 is a block diagram showing the structure of a control system for a polyolefine resin production plant according to a second embodiment of this

invention;
FIG. 8 is a diagram for explaining the concept of a model predictive control in the system in FIG. 7;
FIG. 9 is a functional block diagram for explaining the functional structure of a part of the system in FIG. 7;
FIG. 10 is a functional block diagram for explaining the functional structure of another part of the system in FIG. 7;
FIG. 11 is a flowchart for explaining the operation of the system in FIG. 7;
FIG. 12 is a diagram for explaining a control operation for a polymerization transition in the system in FIG. 7; and
FIG. 13 is a diagram for explaining the control operation of the system in FIG. 7.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
Control systems for a polyolefine resin production plant according to preferred embodiments of this invention will now be described with reference to the accompanying drawings.
First Embodiment
A control system for a polyolefine resin production
plant according to the first embodiment of this
invention will now be described with reference to

The control system in FIG. 1 comprises a bus 1, a

process computer 2, an operators station j, multifunction controllers 4, uninterrupted automatic controllers (UACs) 5, a wiring section 6, raw supply valves 7 and a magnetic disk apparatus 8.
In FIG. 1, the process computer 2 and the operators station 3 are coupled to the bus 1. Those process computer 2 and operators station 3 perform the general control of the system.
Further coupled to the bus 1 is the multifunction controller 4s which are controlled by the operators station 3. Control data for the multifunction controllers 4 are transferred to the respective uninterrupted automatic controllers 5 via the multifunction controllers 4, GO that even when the multifunction controleers 4 are disabled, the uninterrupted automatic controllers 5 can perform desired processing.
The raw supply valves 7 for supplying a raw material to the plant are coupled via the wiring section 6 to the associated multifunction controllers 4, which control the opening and closing of the associated raw supply valves 7.
More specific functions of the control system which is realized by the above hardware structure is exemplarily illustrated in FIG. 2,
A name dat^ base 9 and an operational pattern table 10 shown in FIG. 2 are stored in storage areas set in

the magnetic disk apparatus 8 as an external memory unit of the process computer 2 shown in FIG. 1.
The names of about 100 types of resins, for example, are registered as indexes or index data in the name data base 9. The name data base 9 is a kind of an operation condition data base of each product type. For each type of resin, the specified values of plant operation process data such as the flow rate, temperature and pressure, and the target value for the product like MFR are registered in, for example, a sequential form in the name data base 9.
Operational patterns each consisting of the operation values of a raw-material supply pattern are registered in the operational pattern table 10 each in association with the combination of the target value for a resin type and the target value for the next target resin to be changed. Each operational pattern is equivalent to the combination pattern of a composition value and a raw-material supply pattern.
The process computer 2 compares current operation process data or the operation process data of the type of the current resin with the specified value of the operation process data of the type of the next resin, reads the optimal operational pattern from the operational pattern table 10 based on the combination of the compared operation process data, and controls the multifunction controllers 4 based on the operational

pattern. When the end time for the current rype or resin has been set previously or the like, it is desirable that the process computer 2 should select the optimal operational pattern from the operational pattern table 10 and should select the start time for the optimal operational pattern for each selected operational pattern based on the end time for the current type of resin.
The raw-material supply pattern used in this system will now be explained with the pattern of the feed rate hydrogen taken as an example.
FIG. 3 shows the relationship between the hydrogen concentration (%) in a polymerization reactor and the activity of a polymerization catalyst (relative value) or a change in the activity of a polymerization catalyst with respect to a change in the hydrogen concentration in the polymerization reactor. The trend shown in FIG. 3 is true of the solution polymerization, slurry polymerization, bulk polymerization and gas-phase polrmerization.
In FIG. 3, to reduce the molecular weight at the time of changing one type of resin to another, for example, it should be efficient to feed the sufficient rate of hydrogen at a time until a polymer with a low molecular weight is produced. As is apparent from FIG. 3, however, as the hydrogen concentration increases, the activity of the polymerization catalyst

increases and such a rapid rise in hydrogen concentration causes abnormal polymerization. It is theriefore necessary to compare the hydrogen concentration at the operation value of the current type of resin with the target hydrogen concentration for the next type of resin to select the optimal supply pattern.
According to this system, the hydrogen concentration in the reactor is previously divided into a plurality of zones based on the relationship between the hydrogen concentration and the activity of a polymerization catalyst, and the optimal operational pattern obtained for each segmented zone is registered in the operational pattern table 10.
FIGS, 4A through 4D show the patterns of a change in the supply of a raw material or the patterns of a change in the feed rate of hydrogen for the individual segmented zones, which are to be registered in the operational pattern table 10.
FIG. 4B shows the pattern which increases the feed rate of hydrogen in ancordance with the condition for the next type of resin and is suitable in the case where the difference between the hydrogen concentration in the operation value for the current resin type and that in the target value for the next resin type is less than 0.5%.
The pattern shown in PrtS. 4C increases the feed rate of hydrogen in two steps and is suitable in the

case where the hydrogen concentration in the operation value for the current resin type is less than 2% and the difference between the hydrogen concentration in the operation value for the current resin type and that in the target value for the next resin type is less than 0.5%.
FIG. 4A shows the pattern which temporarily overfeeds the feed rate of hydrogen and is suitable in the case where the hydrogen concentration in the operation value for the current resin type lies off the range illustrated in TABLE 1.
FIG. 4D shows the pattern for the case of reducing the feed rate of hydrogen.
TABLE 1 shows what patterns should be selected based on the combination of the operation value for the current resin type and the target value for the next resin type, the rightmost column showing the feed rate of hydrogen per time.

First Experimental bxampxe The type changing operation for changing one type of resin to another in the polymerization reactor was executed in the case where the hydrogen concentration in the operation value for the current type of resin was less than 2% and the hydrogen concentration in the operation value for the next type of resin was 4% or above.
The process computer 2 read the operation value for the current type of resin from the distributed multifunction controllers 4, read the name of the next type of resin to be changed from the name data base 9, and derived the difference between the target values of the hydrogen concentration foi" both resins. As a result, the pattern in FIG. 4C was selected based on the TABLE 1.
FIG. 4C shows the two-step hydrogen supply pattern in which the feed rate of hydrogen is increased to 20 Nm^/Hr and the feed rate of the catalyst is reduced at the first step, causing the polymerization amount in the polymerization reactor and thus reducing the catalyst concentration, and then the feed rate of hydrogen is further increased to 55 Nm^/Hr. When the hydrogen concentration reaches the target value, the feed rate of hydifogen is reduced to the specified value of 3.5 Nm^/Hr, after which the feed rate of hydrogen is

adjusted to provide the target hydrogen concentration.
It took 2 hours and 10 minutes for the physical property value of the produced polymer to meet the specified value under this control.
First Comparative Example
The same control as done in the first experimental example was executed with an operator manually operating only the distributed multifunction controllers.
At the time of changing one type of resin to another, the operator simultaneously changed the set values of other multiple control loops and monitored those control loops, and performed the condition changing operation while discriminating the set values from the global point of view. Because the set values should be changed in full consideration of safety, the values were set to the minimum levels from the viewpoint of safety.
In this case, to prevent abnormal polymerization, hydrogen was supplied for two hours at the rate of 20 NmVHr and the feed rate of hydrogen was then increased to 40 Nm^/Hr while monitoring the rise in hydrogen concentration. After confirming that the hydrogen concentration rose to the specified level, the feed rate of hydrogen was dropped to 3.5 Nm^/Hr.
It took 3 hours and 45 minutes under this control _for the proper change of the type'of a resin or-for the physical property value of the product to meet the

specified value.
FIG. 5 shows changes in hydrogen concentration (%) of the first experimental example and the first comparative example.
It is apparent from FIG. 5 that this invention can shorten the time for the physical property value of the next type of resin to reach the specified value. Second Experimental Example
A product with a lower molecular weight was changed to a product with a higher molecular weight in the same system structure as used for the first experimental example.
To lower the hydrogen concentration, it is effective to temporarily stop supplying hydrogen and perform dehydrogenation. The "dehydrogenation" means to perform degassing from the polymerization reactor to discharge hydrogen together with the reactive monomer gas. This method can considerably shorten the time needed to reduce the hydrogen concentration. As this method drops the pressure in the reactor which may affect the reaction, however, a large amount of hydrogen should not be discharged. In this respect, this should be used as an auxiliary scheme.
In this experimental example, the process computer 2 selected the pattern in FIG. 4D, temporarily stopped the supply of hydrogen and performed dehydrogenation, and set the feed rate of hydrogen back to the specified

value of 4.5 Nm^/Hr when the hydrogen concentration became the specified level. In this case, the changing of resins from one type to another took 3 hours-
Second Comparative Example
The operator manually performed the same control as done in the second experimental example.
Because the operator should change the set values of other multiple control loops and monitor those control loops at the time of changing one type of resin to another in this case too, the operator performed only the operation to temporarily stop the supply of hydrogen first, and then returned the feed rate of hydrogen to the specified value of 4.5 Nm^/Hr.
It took 5.5 hours to complete the proper change of the type of a I'esin in this case.
FIG. 6 shows successive changes in hydrogen concentration (%) of the second experimental example and the second comparative example. It is apparent from FIG. 6 that this invention can shorten the time for the physical property value of the next type of resin to reach the specified value.
The system shown in FIG. 1 may be realized by connecting a process computer to the bus of a conventional distributed control system so that this process computer accomplish the type change control.
In-executing the type change control, the operation process data of the current type of resin is compared

with the specified value of the operation process data of the next type of resin and the difference between both values is obtained to select the proper operational pattern.
Operational patterns to be registered in the operational pattern table are what is need for the operation process data of the plant, after changed, to reach the specified value of the operation process data of the next type of resin to be produced in order to ensure the continuous transition to the operation for the next type of resin without stopping the operation even after the completion of the operation for the current type of resin. They are considered the optimal operational patterns in that the quantity of products which do not meet the standards and are produced at the time of changing the types of resins and in that the transition to the operation of producing the next type of resin is accomplished safely and efficiently. The operational patterns are obtained by dividing the conditions for changing operation process data into several types of patterns.
As the target value of each type of resin and a successive change in the feed rate of each raw material are registered in the operational pattern table, it is possible to register the operational pattern which is based on -the result af—the analysis previously done through the manual operation for each type of resin.

This can ensure the control that does not depend on the skill levels of the individual operators.
Second Embodiment
A control system for a polyolefine resin production plant according to the second embodiment of this invention will be described below.
The second embodiment is adapted to a continuous polymerization system for polymers which automatically changes polymers from one type to another in the continuous polymerization.
Polymers or high-molecular materials are produced by polymerization. Polymers can be produced by continuous polymerization in the polymerization reactor, This continuous polymerization sequentially produces plural types of polymers which have different molecular weights and different densities.
The switching of the composition in the polymerization reactor should be executed promptly and stably in the continuous polymerization in order to reduce the number of transition products (the off-standards products which are produced transiently e the time of switching from one type of polymer to another) as much as possible when one polymer is changt to another.
For example, the ratio of feed rates of supplied gases, such as the raJ;io of the feed rates of hydrogen ethylene and propylene in changing the molecular weigh

or the ratio of the feed rates of propylene and ethylene in changing the molecular weight/density corresponds to a change in composition in the polymerization reactor and becomes an index for changing the polymer type or the transition.
There is one conventional continuous polymerization technique which over-supplies a catalyst, hydrogen or propylene for a certain period of time in order to cause the ratio of the value measured by a gas chromatography (hereinafter called "gas chromatographic ratio") to approach the target value for the type of a transition product by adjusting the feed rate of the catalyst, the feed rate of hydrogen, the feed rate of propylene or the like with the feed rate of ethylene kept constant, and then returns that ratio to said target value.
As already mentioned, conventionally, switching the polymer type is manually performed by an operator and thus greatly depends on the knowledge- and experience-based skill of that operator. Accordingly, the time for the transition varies depending on operators, thus making it difficult to stabilize the qualities of products around the transition for some operators.
In this respect, the second embodiment of this invention provides a continuous polymerization system for polymers which is capable of automatically and quickly switches one type of polymer to another in the

continuous polymerization while improving the qualities of the products. That is, the second embodiment is a continuous polymerization system for polymers to which the control system of this invention for a polyolefine resin production plant is applied. FIG. 7 shows the hardware structure of the second embodiment.
The continuous polymerization system for polymers sequentially produces different types of polymers through the continuous polymerization which is accomplished by continuously switching compositions in the polymerization reactor.
The continuous polymerization system for polymers in FIG. 7 comprises a model algorithm control section (hereinafter called "MAC section") 21, a distributed control system section (hereinafter called "DCS section") 23, and a polymerization reactor 28.
The MAC section 21 sets a model pattern for the DCS section 23 and performs fuzzy control. Accordingly, the MAC section 21 manages model pattern parameters fuzzy parameters and a name condition file and manages the fuzzy control.
The DCS section 23 executes model-pattern based program control to perform the general control of the system and the sequence control.
The MAC section 21 executes the model predictive control using a model pattern- obtained by patterning the overaction part in a test sequence prepared by the DCS

section 23.
The "model pattern" is the operational pattern of the feed rates of hydrogen and propylene set on the basis of the past achievements for the individual polymer types. A scheme of patterning the transitional operations based on the experiences of skillful operators to construct models for the individual polymer types belongs to the model predictive control. For the optimization purpose, a control apparatus which employs the model predictive control should be used.
The model predictive control includes MAC (Model Algorithmic Control) and DMC (Dynamic Matrix Control), for example. The typical method for the model predictive control uses the dynamic characteristic model of the process in the control algorithm to predict the future behaviors Yp(t), Yp(t+1), ... of a control target from the previous operational data and to control it within the finite time, as illustrated in FIG. 8.
With the use of the model pattern alone, the fuzzy control in the MAC section 21 cannot cope with the control of the physical property which changes from time to time. Therefore, the fuzzy control is used to control the physical property to the target value. In the fuzzy control, membership functions and fuzzy rules are set in association with the manual control operation so as to achieve a human-like control operation.
The continuous polymerization system for polymers

performs a fuzzy operation on the process feedback values and reflects the operation results on the process again. More specifically, the continuous polymerization system for polymers executes sequence control which performs control based on model patterns attained from the past operational achievement data for the individual polymer types and indirect advanced control which performs a fuzzy operation at a higher level than the sequence control and sets the operation results in the DCS section 23 for advanced control.
The DCS section 23 may be constructed by using a general-purpose computer system (e.g., "CENTUM" (product of YOKOGAWA ELECTRIC CO.) or "TDCS" (product of YAMATAKE HONEYWELL CO.)),, which includes an operators station 24, an operators console 25 and a field controller 26.
The field controller 26 has a batch setter 55 to be discussed later.
The MAC section 21 has a hard disk drive, a random access memory and an RS71 interface card, though not illustrated, and operates the system based on software. The software includes a line controller, BASIC interpreter, an interface card utility, an interactive graphics utility, a trend library, a data base utility, DOS (Disk Operating System), a conversion utility, a shell and a fuzzy processing program (such as "AdMAS" which is a packaged program).
A catalyst, hydrogen, ethylene and propylene

Dnomer or the like are properly supplied to the Dlymerization reactor 28 for polymerization. The field ontroller 26 receives data such as the feed rate of the atalyst, the feed rate of hydrogen, the feed rate of ropylene and the like. The field controller 26 ontrols the pressure and the temperature in the olymerization reactor 28, and the gas chromatographic atio.
The polymerization reactor 28 performs continuous lolymerization. The polymer may be polyolefine which is i high-molecule compound obtained by the polymerization )f a-olefines each of which has a carbon-carbon double )ond in the molecules. Polyolefine includes low-density )olyethylene which is synthesized by radical polymerization. The polyolefine may also be polyethylene having an intermediate or high density or a polyolefine wax.
The functional structure of the continuous polymerization system for polymers that can be achieved by software will now be described.
FIG. 9 shows the functional structure of the MAC section 21, and FIG. 10 shows the functional structure of the DCS section 23.
The software of the continuous polymerization system for polymers consists of the MAC function portior shown-in FIG. 9 and the DCS sequence function portion also shown in FIG. 9. The MAC function portion include!

a data base 29, a fuzzy control section 31, a trend monitor section 32, a data base input/output section 33, a data collecting section 36, a data setting section 37, a fuzzy processor 38, a simple simulator 41, a transition data collecting section 42, a transition setting section 43, a file selector 44 and a data analysis/trend section 45.
The fuzzy control section 31 manages the data input/output of the fuzzy processor 38 and name condition parameters, and performs fuzzy control like data processing and operation and pressure control of model patterns. In other words, the fuzzy control section 31 executes fuzzy control in such a way that sub-parameters like the gas chi-omatographic ratio, pressure and temperature, which change in accordance with the values of control parameters like the feed rate of hydrogen and the feed rate of a catalyst, are permitted to approach the target values associated with the next type of polymer to be changed, upon passage of a predetermined time passes after the values of the control parameters change.
The trend monitor section 32 presents the trend display of the transition status in real time to displa; the contents and status of the fuzzy control-
The data base input/output section 33 manages the input/output access to a data file 29A in the data base 29, such as the aforementioned model pattern parameters

and name operation condition data, and inputs and changes those data.
The simple simulator 41 has a process simulation function to store a simple model equation and a processing model equation of the process simulation, infer the physical property of a polymer and inspect the fuzzy output.
The data collecting section 36 collects data from the DCS section 23 through the communications software that is invoked on the line controller and stores the data in a common area. That is, the data collecting section 36 serves as a detecting section to detect the values of the aforementioned sub-parameters.
The fuzzy processor 38 incorporates a fuzzy operation section which is provided as a subprogram of the fuzzy processing program, and executes a fuzzy operation under the control of the fuzzy control section 31. That is, the values o£ the aforementioned control parameters are attained from the fuzzy operation on the values of the sub-parameters that are collected by the data collecting section 36.
The data setting section 37 sets the model pattern data of the type of a resin and the values obtained through the fuzzy operation into the batch setter 55 of the field controller 26.
The fuzzy processing progr.ajn builds fuzzy rules, membership functions and the like. Each fuzzy rule has

a condition (IF) part and a conclusion (THEN) part each consisting of a fuzzy set.
The following are some examples of fuzzy rules. Rule 1
IF X is Aj and y is Bj THEN z is Cj Rule 2
IF X is A2 and y is B2 THEN z is Cj
Both rules have two condition part variables "x is A" and "y is B" and one conclusion part variable "z is C." In the rules, Aj^, Aj, B^, B2, C^ and C2 are fuzzy sets.
The membership functions define the parts of a fuzzy rule for quantization.
The fuzzy operation obtains the matchability (goodness of fit) of the condition part of each rule with given values, obtains the inference result of each rule based on the attained matchability, and then attains the final inference result from the inference result of each rule.
The simple simulator 41 inputs data from the transition data file 29A to the fuzzy processor 38, provides the trend display of the output values on the trend monitor section 32 and adjusts the setting of the membership functions and fuzzy rules for the fuzzy control based on the displayed results.
The transition data collecting section 42 stores, for example, 88 pieces of data from one hour before th
beginning of the transition up to the end of the transition into the data file 29A in the data base 29. The transition setting section 43 sets the trend name and the time for starting the data collection.
The file selector 44 selects the transition data file 29A to be displayed. The data analysis/trend section 45 presents the trend display of the transition data of the designated name in the data file 29A.
Then, the DCS sequence section 50 has a system control sequence section 51, a parameter setting sequence section 52, a fuzzy control sequence section 53, a model pattern sequence section 54 and the aforementioned batch setter 55.
The fuzzy control sequence section 53 and model pattern sequence section 54 control the polymerization reactor 28.
The system control sequence section 51 controls the selection of the resin type, the selection of the sequence in use, the start and stop of the sequence, the selection of a model pattern parameter and the communication with the MAC section 21. The resin type is managed by an 8-bit code; a binary code is handled with a type number in a decimal notatioh-
The parameter setting sequence section 52 communicates with the MAC section 21 upon the initiation of the sequence after a resin type is selected, and sets model pattern parameters in the batch setter 55 of the

DCS section 23.
The parameter setting sequence section 52 sets the same value as two model pattern parameters in the batch setter 55. This sequence section 52 compares the two values, set in the batch setter 55, with each other and determines that the proper communication has been accomplished when those two values are the same, and sets this value in a program setter (not shown). When the two model pattern parameters in the batch setter 55 differ from each other, the sequence section 52 sends a communication error message to the MAC section 21 and retries the communication.
The fuzzy control sequence section 53 manages the fuzzy operation values which are to be set by the data setting section 37 of the MAC section 21 and causes the batch setter 55 to write in a control tag- This sequence section 53 is enabled when a predetermined period of time passes after the setting of the model pattern parameters, and controls the communication between the MAC section 21 and the DCS section 23.
The model pattern sequence section 54 is a patterning sequence which optimizes the patterning and constructs a model pattern for the resin types whose fuzzy parameters have not been built yet. This sequence section 54 changes the values of control parameters based on the moder^attern for each type at the time of switching one type of polymer to another in the

continuous polymerization.
The control operation for the continuous polymerization in the continuous polymerization system for polymers will now be discussed.
FIG. 11 is a flowchart illustrating the continuous polymerization control, and FIG. 12 illustrates the control characteristic of the polymerization transition.
First, the MAC section 21 builds model patterns (step SOO). In this case, the test pattern sequence performed patterning for 18 types, for example. Further, the patterning was performed with respect to 3 parameters: the feed rate of hydrogen, the feed rate of a catalyst and the feed rate of propylene.
FIG. 13 exemplifies a model pattern for hydrogen. As shown in FIG. 13, the hydrogen pattern indicates the hydrogen supply of 30 Nm^/Hr for 25 minutes and the hydrogen supply of 120 Nm^/Hr for 30 minutes at the time of overaction.
With regard to the catalyst, the activity always varies, making it difficult to previously determine the feed rate. Therefore, the ratio of a change in activity is factorized based on the past achievements, and the feed rate is computed by the MAC section 21 spontaneously upon the activity at the transition time. The computed feed rate is set in the DCS section 23. _ Because the constructed model pattern has been-optimized to a certain degree from the past

aohievements, the polymerization transition time can be shortened. For instance, the time needed for the transition work (including feed-down and feed-up) is about 12 hours on average. The automation system for transition polymerization shortens this transition time to 10 hours, two hours shorter.
The setting of the transition type, and the start and stop of the sequence are executed on the screen of the operators station 24 and a message is sent to the MAC section 21 from the DCS section 23 (step SOI),
The MAC section 21 receives the message from the DCS section 23 and sends an acknowledge message to the DCS section 23 (step S02).
Next, the MAC section 21 determines if the message from the DCS section 23 is corx-ect (step 503). VJhen the message from the DCS section 23 is correct, the data setting section 37 in the MAC section 21 sets model pattern parameters in the batch setter 55 in the DCS section 23 (step 304). In this example, two same model pattern parameters are set in the batch setter 55.
Then, the DCS section 23 determines if the set model pattern parameters have been transferred by comparing a flag with the two same model pattern parameters (step S05),
When the model pattern parameters have been transferred properly, it is determined if fuzzy setting of the type has been performed (step S06). When this

fuzzy setting has been done, the system control sequence section 51 selects and invokes the fuzzy control sequence section 53 (step S07).
When the fuzzy control sequence section 53 is selected, the transition operation is initiated with the previously defined model pattern and is switched to fuzzy control after a predetermined period of time (step SOS).
More specifically, overaction is performed to change the feed condition to the condition for the next type of product in the initial state at the time of transition, and after the passage of the predetermined period of time, fuzzy control is performed to set the gas flow rate to the H/E target value shown in FIG. 12.
The "predetermined period of time" is the time for setting the control target value to the value for the next type of product upon the execution of the overaction after the initiation of the transition, and is set for each type.
In this case, the fuzzy control section 31 executes the gas chromatography control (H/E control) and pressure control based on the control algorithm during the fuzzy control period as shown in FIG. 12. The control algorithm can be changed for fine adjustment even during operation. The control cycle is separated to two paxts for the gas chromatography control and pressure control, each of which can be set arbitrarily.

For example, the chromatography control is carried out every 3 minutes while the pressure control is executed every minute.
When the physical property reaches the target valu( and becomes stable, the transition is ended and the normal operation is initiated (step S09). The fuzzy control continues even in the normal operation to perform stabilization control in this operation.
The fuzzy control sequence continues until the feed-up, the feed-down and the next transition. When the fuzzy control sequence is stopped arbitrarily based on the control table, control is switched to the ordinary control by the manual operation.
When fuzzy setting has not bee made, the system control sequence section 51 selects and invokes the model pattern sequence section 54 (step SIO).
This sequence section 54 starts the model pattern control shown in FIG. 12 at the time of overaction (ste Sll). In the pattern compensation in this case, the pressure of n minutes later is inferred for the pressui control,
Since overaction is taken in the predetermined period of time, the physical property is not controllec and only the pressure in the polymerization reactor is controlled.
The data collecting section 36 of *he high-rank MJ section 21 collects, for example, five pieces of

pressure data each obtained every minure ana ou-cputs xne ratio of change in the collected pressure data (step S12).
Pressure change for one minute:
dpl=P-Pl, dp2=Pl-P2, dp3=P2-P3,
dp4=P3-P4, dp5=P5-P6
Then, the data collecting section 36 infers pressure of one minute later from the current point, based on the data on the ratio of change (step S13).
Inferred pressure after one minute:
AVPA=SUM(dpl, dp2, dp3, dp4, dp5)/5 +reactor pressure D201PS
The data collecting section 36 determines if the inferred pressure value has reached the pressure point (upper limit/lower limit) previously determined by the type condition (step S14).
At this point of time, the data collecting section 36 changes the amount of the catalyst to control the pressure (step S15). The amount of a single operation of the feed rate of the catalyst was the finite difference between the first point and the second point in the model pattern based on the past achievements.
When the pressure reaches the point determined by the type condition, the ^4AC section 21 outputs an upper-limit pressure signal and a lower-limit pressure signal to the DCS sectio^n 23 to perform the sequence-based control (step S16).

Since the gas chromatography control is not executed during this period, the manual operation is necessary when the operation enters the stable state. The pattern sequence is used to optimize the model pattern, not for automation.
The automation of the polymerization transition i carried out by the construction of the model patterns and the fuzzy control. The pattern of the test sequen which has constructed the model patterns under the model-pattern based program control can be automatical set by the MAC section 21 to be used for the optimization of the type-by-type patterns.
It is therefore possible to develop the scheme oi constructing the fuzzy parameters for the automatic switching of polymers as needed, even for other types polymers than the typical polymers which are automatically changed from one to another.
During the predetermined period of time, only th The system may also use a feed rate of at least of a-olefines having 2-8 carbon atoms, e.g., jethylene propylene, butylene and the like, as the control parameter.
It is apparent that, in this invention, a wide

range of different working modes can be formed based on the invention without deviating from the spirit and scope of the invention.

WE CLAIM;
1 . A control apparatus for a polyolefine resin
production plant comprising:
a production system capable of producing plural types of polyolefine resins;
name data base means for previously registering names of a plurality of polyolefine resins producible by said production system as index data;
pattern storage means for previously storing an operational pattern for an optimal type changing operation as an operational pattern table for each combination of a name of a current resin under production in a currently running production process by said production system and a name of a next target resin to be produced by said production system;
pattern selecting means for comparing said name of said current resin with said name of said next target resin to discriminate a combination thereof and selecting an operational pattern corresponding to a discrimination result from said operational pattern table; and
change control means for executing process control according to said operational pattern selected by said pattern selecting means to change a type of a resin to be produced by said production system.
2 . The control apparatus according to claim 1,
wherein said pattern storage means has means for

storing an operation pattern having a control target value for each resin type and information on a successive change in a feed rate of a raw material in said operational pattern table.
3. The control apparatus according to claim 2, wherein said pattern selecting
means has means for setting a start time for control by said selected operational
pattern for a type changing operation to an optimal time for each operational
pattern based on a given production end time for a type of a current resin.
4. The control apparatus according to claim 2, wherein production system
has a polymerization reactor to produce different types of polyolefine resins
having different molecular weight and a copolymer composition of polyolefine;
and said pattern storage means to register an optimal operational pattern given
for each of a plurality of zones, obtained by previously dividing a hydrogen
concentration in said polymerization reactor based on at least a relationship
between a hydrogen concentration and activity of a polymerization catalyst, into
said operational pattern table.
5. A control apparatus for a polyolefine resin production plant, substantially
as herein described with reference to the accompanying drawings.

Documents:

1624-mas-1995 abstract.pdf

1624-mas-1995 claims.pdf

1624-mas-1995 correspondence-others.pdf

1624-mas-1995 correspondence-po.pdf

1624-mas-1995 description(complete).pdf

1624-mas-1995 drawings.pdf

1624-mas-1995 form-1.pdf

1624-mas-1995 form-26.pdf

1624-mas-1995 form-29.pdf

1624-mas-1995 form-4.pdf

1624-mas-1995 form-9.pdf

1624-mas-1995 petition.pdf

« Previous Patent

Next Patent »

Patent Number

193134

Indian Patent Application Number

1624/MAS/1995

PG Journal Number

35/2005

Publication Date

16-Sep-2005

Grant Date

24-May-2005

Date of Filing

11-Dec-1995

Name of Patentee

MITSUI PETROCHEMICAL INDUSTRIES, LTD

Applicant Address

2-5 KASUMIGASEKI 3-CHOME, CHIYODA-KU, TOKYO

Inventors:

#	Inventor's Name	Inventor's Address
1	SHIGEKI HAYASHI	C/O MITSUI PETROCHEMICAL INDUSTRIES, LTD., 3 CHIGUSA KAIGAN, ICHIHARA-SHI, CHIBA-KEN
2	MICHINORI TAYAMA	C/O MITSUI PETROCHEMICAL INDUSTRIES, LTD., 3 CHIGUSA KAIGAN, ICHIHARA-SHI, CHIBA-KEN
3	HIROYUKI MIZUOCHI	C/O MITSUI PETROCHEMICAL INDUSTRIES, LTD., 3 CHIGUSA KAIGAN, ICHIHARA-SHI, CHIBA-KEN
4	HIROKAZU KATSUYAMA	C/O MITSUI PETROCHEMICAL INDUSTRIES, LTD., 3 CHIGUSA KAIGAN, ICHIHARA-SHI, CHIBA-KEN

PCT International Classification Number

C08F2/00

PCT International Application Number

N/A

PCT International Filing date

PCT Conventions:

#	PCT Application Number	Date of Convention	Priority Country
1			NA