Title of Invention

A METHOD FOR FILTERING A VARIABLE

Abstract

A device and a method for filtering a variable are iescribed. A first filter means is used to form an output variable as a function of an input variable, the first filter means having at least a delaying effect. The input variable of the first filter means is corrected using a correction variable which is acquired on the basis of the input variable of the first filter means by filtering using a second filter means. (Figure 2) 1

Full Text

11.05.00 Bg/Hx ROBERT BOSCH GMBH, 70442 Stuttgart Method and device for filtering a signal Prior Art The invention relates to a method and a device for filtering a signal according to the preambles of the independent claims - A method and a device for filtering a signal is [sic] known, for example, from DE 195 37 787. In said publication the driver's request quantity is filtered by means of an I/O module. The filtering means is embodied in such a way that, for example, rapid changes in driver quantity requests (pedal value) do not have an undamped effect on the metering of fuel and the excitation of longitudinal oscillations of the vehicle is thus avoided. Such filtering means in order to damp the excitation of systems have [sic] the disadvantage that they [sic] generate a drag error when there is a ramp-like change in the input variable. That is to say the output variable follows the input variable only with a delay. This has the effect of reducing the drive torque, for example in an application in an internal combustion engine. Advantages of the Invention The procedure according to the invention provides the advantage that corresponding drag errors can be compensated without having to cope with restrictions in the filter effect, in particular when there are abrupt changes in the input variable. Advantageous and expedient refinements and developments of the invention are characterized in the subclaims. Drawing The invention is explained below with reference to the embodiments illustrated in the drawing. Figure 1 shows the basic design of a fuel metering system, and Figure 2 shows a block diagram of the procedure according to the invention. The invention is illustrated below using the example of a fuel quantity signal in an auto ignition internal combustion engine. The invention is not restricted to this application. It can also be used with other signals, in particular with signals which are used for controlling internal combustion engines. The method is in particular suitable for signals which influence or characterize the torque which has been output. Such signals are, for example, a fuel quantity signal, signals for actuating power-influencing actuators, a quantity request signal, the output signal of an accelerator pedal sensor or a rotational speed signal. Figure 1 shows the basic design of a fuel metering system of an internal combustion engine. 10 designates an accelerator pedal position sensor and 11 designates a rotational speed sensor. A setpoint value controller 12 is connected to the accelerator pedal sensor and to the rotational speed sensor 11. The output signal MEW of the setpoint value controller which corresponds to the driver request quantity is fed to an I/O module 13. The rotational speed signal N of the rotational speed sensor 11 is fed to an interference variable regulator 14. The output signal MEF of the I/O module 13 and the output signal MES of the interference regulator 14 are superimposed at another point and form the quantity signal MEA which is fed to an actuator device 15. An corresponding fuel quantity is metered to the internal combustion engine (not illustrated) as a function of the signal MEA. On the basis of the accelerator pedal position of the rotational speed, the setpoint value controller 12 calculates the driver request quantity MEW which is necessary to make available the driving power desired by the driver. In systems without judder suppression, this signal is fed directly to the actuator device 15. The actuator device 15 converts this signal into an actuation signal to be applied to the corresponding actuator elements. Thus, for example in series-mounted pumps there is provision for an actuator control circuit to adjust the control rod position to a corresponding value. In time-controlled systems, the actuator device 15 outputs an actuation signal for a quantity-determining solenoid valve or a piezoelectric actuator. In order to be able to compensate judder vibrations, there is provision for the driver request signal MEW. to be filtered using an I/O module 13. The I/O module 13 has at least a delaying effect. Thus, for example, filters with first-order time delay behavior can be used. It is particularly advantageous if filters which do not have any other components are used as I/O modules. Furthermore, the rotational speed signal N is fed to an interference regulator 14. The new method of operation of this device is described in DE 195 37 787. If the filter 13 which forms the I/O module has at least delaying behavior, for example a Tl element, a drag error occurs when there are specific changes in the input variable of the filter 13. That is to say the output variable follows the input variable only with a delay. According to the invention, this drag error is eliminated in that a correction value, which is formed on the basis of the input variable, is applied to the input of the filter. For this purpose, the input variable is preferably derived according to time, i.e. in a differentiated fashion and then weighted with a value which can in particular be predefined. This weighting factor is preferably predefined as a function of the transmission behavior of the filter to be corrected. Here, the time derivative of the input variable is limited in order to retain the filter ^ effect when there is a rapidly changing input variable despite the measures to counteract drag errors. Figure 2 is a detailed illustration of the I/O module with such a correction. Elements which have already been described in Figure 1 are designated with the corresponding references. The actual filter of the I/O module is designated as a first filter 100. The input variable MEW of the I/O module 13 is fed, on the one hand, with a positive sign to a logic-linking point 125 and, on the other hand, to a second filter 110. The output signal of the logic-linking point 125 is fed to the first filter 100. The output signal of the second filter 110 is fed to a second logic-linking point 115 via a limiter 112. The output signal of the logic-linking point 115 is preferably fed with a positive sign to the logic-linking point 125. The output signal of a factor predefining means 12 0 is present at the second input of the second logic-linking point 115. The output signal of the first filter 100 forms the output variable MEF. In one refinement it is also possible to provide for the limiter 112 to be arranged downstream of the logic-linking point 115. This means that the limiter 112 limits the correction variable with which the input variable of the first filter 100 is corrected at the logic-linking point 125. A particularly advantageous refinement of the procedure according to the invention is represented by dot-dashed lines. In said procedure, the input variable is additionally fed via an amplifier 140 to a logic-linking point 130 at whose second input the output variable of the first filter 100 is present. These two variables when logically linked then form the output variable MEF. The second filter 110 is preferably embodied as a differentiator. The second filter 110 comprises at least a differentiating component. For example, the second filter can also be embodied as a PD element or as a DT element- The output variable of the second filter 110 is limited by the limiter 112 to the maximum permissible absolute values, in order to ensure the filter effect when there are rapid, in particular abrupt, changes in the input variable MEW. The limiter 112 is dimensioned in such a way that the limitation is ineffective when the input variable changes slowly, and such that the filter 110 makes an uninfluenced contribution to the correction of the input variable of the first filter 100. When there are slow changes in the input variable, the second filter 120 has a relatively large influence on the filtered variable. As a result, according to the invention the drag error is avoided. When there are abrupt, that is to say rapid, changes in the input variable, the limitation is effective, as a result of which the corresponding contribution of the second filter 110 to the correction of the input variable of the first filter is only small. When there are rapid changes in the input variable, the second filter 120 has a relatively small influence on the filtered variable. In this case, the first filter 100 has a large influence on the filtered variable. The output signal of the second filter 110 is weighted with a predefinable weighting factor of the factor predefining means 120 at the logic-linking point 115. The weighting factor can be predefined in particular as a function of the transmission ratio of the first filter 100. In one preferred embodiment, the first filter 100 has the transition function: K/(T*s+1) Here, the variable T is usually referred to as a delay time constant, and the variable K is usually referred to as a proportional amplification. The factor of the factor predefining means 120 is preferably identical to the time constant T. This means that the output signal of the second filter 110 which is limited by the limiter 112 is weighted with the factor of the factor predefining means 120, i.e. with the delay time constant T of the first filter 100. In the second, particularly advantageous, embodiment, the amplifier 140 has the gain factor V. The proportional amplification K of the first filter then assumes the value K = 1 - V. According to the invention, the input variable MEW of the first filter 100 is corrected as a function of the the [sic] input variable MEW of the first filter 100. This means that a correction variable is determined on the basis of the input variable MEW of the first filter in order to correct this input variable. In a simple embodiment, the input variable is derived over time or differentiated and then weighted with a factor. The factor is determnined here essentially by means of the transmission behavior of the first filter. The factor preferably corresponds to the delay time constant T of the first filter. It is particularly advantageous if only part of the signal is corrected. This is carried out in that the proportional amplification K of the first filter is selected to be smaller than 1 and a correspondingly amplified input signal is fed to the input signal of the first filter. aim 1. Device for filtering a property using an initial (first) filtering agent or material for obtaining an output parameter, depending upon the input parameter of the filtering agent, where the initial filtering agent possesses at least a retarding effect is thereby characterized that, the input parameter of the initial filtering agent is corrected by means of a correction parameter. Such a connection parameter is obtained, proceeding from the input parameter of the first filtering agent, through filtering by using a second filtering agent. 2. The device as per Claim 1 above is thereby characterized that the second filtering agent possesses at least a differentiated behaviour. 3. The device as per Claim 1 or 2 above is thereby characterized that an output parameter of the second filtering agent is measured with a Factor. 4. Device according to claims mentioned above, is thereby characterized that the output parameter of the second filtering agent or the correction parameter is limited. 5. Device according to Claim 3 is thereby characterized that the factor is dependent on the transfer behaviour of the first filtering agent. 6. Device is thereby characterized that in addition the output parameter of the first filtering agent together with the measured input parameter of the first filtering agent can be corrected. 7. Process for filtering of a property or parameter with a first filtering agent for obtaining an output parameter, depending on an input parameter of the filtering agent, where the first filtering agent possess at least a retarding effect, is thereby characterized that the input parameter of the first filtering agent is corrected, using the correction parameter, obtained through filtering through the second filtering agent, proceeding from the input parameter of the first filtering agent. 8, A device for filtering a variable, substantially as hereinabove described and illustrated with reference to the accompanying drawings. 9, A method for filtering a variable, substantially as hereinabove described and illustrated with reference to the accompanying drawings.

Documents:

abs-in-pct-2002-1804-che.jpg

in-pct-2002-1804-che-abstract.pdf

in-pct-2002-1804-che-claims filed.pdf

in-pct-2002-1804-che-claims granted.pdf

in-pct-2002-1804-che-correspondnece-others.pdf

in-pct-2002-1804-che-correspondnece-po.pdf

in-pct-2002-1804-che-description(complete) granted.pdf

in-pct-2002-1804-che-description(complete).pdf

in-pct-2002-1804-che-drawings.pdf

in-pct-2002-1804-che-form 1.pdf

in-pct-2002-1804-che-form 26.pdf

in-pct-2002-1804-che-form 3.pdf

in-pct-2002-1804-che-form 5.pdf

in-pct-2002-1804-che-other documents.pdf

in-pct-2002-1804-che-pct.pdf