Title of Invention

A METHOD FOR OPTIMIZATION OF OPERATING PARAMETERS

Abstract

The invention relates to a method for optimising operational parameters of a wind farm, in particular in relation to the rotor/generator system thereof, by determining an upper and a lower interval threshold value (110) for a parameter which is to be optimised. The method according to the invention is characterised by a cycle of the wind farm having alternating operations (112, 114, 118, 120), provided with interval threshold values. A data set is generated (116, 122) having a target variable, and repeated (126) according to an adjustable number of times. The data set is evaluated in relation to the interval threshold values and a quality level is formed. The interval threshold value is identified by means of the low quality level (130). At least said interval threshold value is modified by displacing about a step value ? in the direction of the other interval threshold value (134, 136). Said cycle is repeated (137). Said inventive method is simple to carry out and enables optimum parameters to be automatically detected even under difficult conditions, for example, stochastically acting disturbances (wind fluctuations). The invention also relates to a wind farm for carrying out said method.

Full Text

Method for optimizing operational parameters on wind farms The invention relates to a method for optimization of operating parameters of wind energy installations, in particular with respect to those of the rotor/generator system, in which an upper and a lower interval limit value are predetermined for the parameter to be optimized. Modern wind energy installations are complex structures which are operated with the aid of a complex control system. The control system is in this case designed to operate as well as possible various wind energy installation systems in accordance with the requirements of the operator of the wind energy installation and the operator of the grid system to which the wind energy installation is connected. In this case, the expression as well as possible generally means that a high energy yield is achieved, although attention must also be paid to the noise emission being as low as possible, to a low load on the drive train, and on the load-bearing structure, and furthermore, to good compatibility with the electrical grid system. The rotor/generator system is particularly important in this context. The rotor together with its blades is of critical importance for the absorption of energy from the wind, that is to say for the conversion of wind energy to mechanical energy, which is in turn converted to electrical energy by the generator. In addition to the design, one important parameter of the rotor relates to the rotor blade length and the profile shape of the rotor blades, the pitch angle which is generally variable. The pitch angle is defined as the angle between a representative profile chord and the rotor plane. The pitch angle is extraordinarily important for definition of the aerodynamic conditions on the rotor blade. It is the governing factor for the flow conditions between the rotor blade and the apparent wind acting on the rotor blade. The apparent wind comprises the true wind and the wind induced by the movement of the rotor blade. The strength and direction of the apparent wind therefore depend on the rotation speed of the rotor, which is in turn dependent on the speed of the true wind and the pitch angle. This therefore results in a closed-loop system for rotor rotation speed control. It is therefore difficult to find optimum values for the pitch angle, as a parameter. Additionally, values have been used as presets for the rotor blade pitch angle determined in advance during operation of wind energy installations. One disadvantage of this method is that it is highly dependent on the accuracy of the initial calculation. Furthermore, this method is sensitive to discrepancies in the rotor blades from the ideal shape for example resulting from unavoidable manufacturing tolerances. This method therefore does not lead to satisfactory overall results. Furthermore, it is also known from obvious prior use for long-term measurements to be carried out on the wind energy installation, and for appropriate correction values to be derived from them. This avoids the above disadvantages, but this method is time-consuming and costly. It is also known (DE-A-101 27 451) for parameter variations to be carried out controlled by a timer on a cy- clic basis. The method is based on the idea of varying one parameter within predetermined limits until an optimum value is found. One disadvantage of this method is that it is susceptible to finding a local optimum, This conceals the risk of a global optimum not being identified. Furthermore, the method has the disadvantage that it cannot distinguish between stochastic fluctuations in the wind strength and wind direction and real improvements resulting from the changed parameter settings. The method therefore does not always provide satisfactory results. The invention is based on the object of improving a method of the type mentioned initially such that good optimization is achieved in conditions which are made more difficult by stochastic fluctuations. The solution according to the invention lies in the features of the independent claims. Advantageous developments are the subject matter of the dependent claims. According to the invention, in the case of a method for optimization of operating parameters of a wind energy installation, in particular with respect to its rotor/generator system, a cycle is predetermined with an upper and a lower interval limit value being defined for a parameter to be optimized, in which cycle the wind energy installation is operated alternately with the interval limit values, with one data record in each case being produced with a target variable, to be precise over a variable number of repetitions, evaluation of the data records relating to the interval limit values with a quality measure being formed, identification of the interval limit value with the poorer quality measure, replacement of at least this interval limit value by shifting through a step value A in the direction of the other interval limit value, and repetition of the cycle. The invention is based on the idea of using an iterative process for parameter optimization. The invention has found that an iterative process with a sufficiently large number of samples makes it possible to compensate for stochastic fluctuations so that they virtually no longer have any disturbing effect on the result. In contrast to the known methods, the statistical method according to the invention is not negatively influenced by stochastically fluctuating variables, such as the wind speed and direction. It is therefore very good in practical application, even in difficult conditions. The concept of the method is of captivating simplicity, and initially requires only two interval limit values for the parameters to be optimized. Building on this, the method is carried out automatically and is able to determine not only an optimum value located between the interval limit values but also an optimum value located outside the interval limit values. The invention achieves this by alternately and successively measuring and storing the desired target parameter relating to the interval limit values. This interaction is repeated a specific number of times. This number can be predetermined as a fixed number or may be variable, depending on the parameter and the optimization quality already achieved, as expressed by the quality measure. Furthermore, the values for the target variable are each evaluated for the two interval limit values, and a quality measure is formed for each of the interval limit values, as a function of the result. The quality jacent wind energy installations typically have a negative influence on the measurements of wind parameters, such as the wind strength or wind direction. Optimum operation of the wind energy installation is therefore virtually impossible- The invention is based on the surprising discovery that there is no need at all to measure these parameters. Instead of this, the respectively optimum pitch angle is determined by means of the method according to the invention in the respectively prevailing wind conditions, irrespective of the wind direction and strength. A large number of samples or iterations can admittedly lead to lengthening of the time that is required to carry out the method, but this is not of major importance in the case of an automated method, as envisaged by the invention. The invention results in better utilization of the wind energy installation. This reduces yield reductions resulting from a non-optimum choice of the parameter to be optimized. A number of expressions used in the following text need to be explained: A parameter is a system coefficient which influences the system behavior. This may be a parameter of the mechanical system or of the electrical system, or a parameter of the control device. The target variable to be optimized is a system variable. This results as a consequence of other variables and parameters. Examples of the target variable, are, inter alia, the electrical power output, the noise level emitted, the structural load on the machine re- suiting from bending and/or vibration loads, as well as variables relating to grid-system compatibility in particular such as flicker. The target variable need not necessarily be a scalar, it can also be composed vecto-rially from a plurality of variables. The interval limit values can be predetermined directly at the start of the method. However, it is also possi- . ble to calculate the interval limit values, to be precise from an operating value (aB) and a scatter value (CXOFF) • The latter allows simple inclusion in an existing operating control system. In this case, the parameters as determined in the conventional manner by the operating control system are each used as a start value for the operating value (aB) . There is then no need to separately determine a suitable value for starting the method according to the invention. In general, both of the interval limit values will be changed at the end of a cycle. This is preferably done by changing the operating value. However, it is also possible to envisage a situation in which the aim is to additionally reduce the interval. The change to the better interval limit value is then carried out to a lesser extent, to be precise reduced by a reduction value 5. The interval is thus reduced cyclically and successively until, in the end, the method according to the invention converges at an optimum value. It is also possible to provide for the reduction value to be negative. The range between the interval limit values then widens. This may be advantageous at the start of the optimization process when the aim is to quickly search through quite a wide interval. The cycle is expediently repeated until a predeter-minable termination criterion e is reached. In this case, e expediently represents a difference magnitude relating to the quality measure. If this magnitude is sufficiently small, then the optimization process can be ended. However, it is not absolutely essential to use the quality measure for the termination criterion. It is also possible to provide for the cycle to be repeated until the interval limit values have reached a separation of only e'. The step width A and, if appropriate, the reduction value 5 are expediently determined by means of a predictor. This has the advantage over rigid presetting that the predictor frequently makes it possible to achieve faster convergence of the method according to the invention towards an optimum value. Optimization processes which are known per se can be used for the predictor, for example genetic algorithms. In one proven embodiment, the quality measure is calculated by addition. This is an evaluation and quality-measure formation process which can be calculated particularly easily and efficiently. If the aim is also to take account of the number of data items used, then an averaging process can also be provided. In this case, the expression an average should be understood in the wide sense; the expression covers not only the arithmetic mean but also other calculation methods, such as the geometric mean. It is self-evident that more complex statistical methods can also be provided, in particular those in which the data is weighted over a pre-determinable time interval (for example by means of a rectangular or Hamming window). It is frequently necessary to optimize only one target variable- However, it may also be necessary to optimize a plurality of target variables. In this case, the quality measure is expediently formed multidimension-ally. This is preferably done in the form of a vector. One parameter for the optimization process is preferably the rotor blade pitch angle and the target variable is the electrical power output. As explained initially, the pitch angle of the rotor blades is a significant criterion for the efficiency of absorption of mechanical energy in order to drive the generator from the wind energy. This is therefore a particularly important fact for the electrical power that is generated, and thus for the wind energy installation yield. On the other hand, however, the pitch angle is governed by the flow conditions between the apparent wind and the individual rotor blade, with the apparent wind in turn being governed by the true wind and the rotation speed of the rotor. The rotation speed of the rotor is in turn predominantly governed by the pitch angle. This therefore results in a closed-loop relationship, in which the optimization process according to the invention can be used particularly advantageously. It is self-evident that the method can also be used for other parameters. By way of example, the generator torque characteristic can be used as a parameter. This likewise has a significant influence on the rotor rotation speed, and thus also on the blade pitch angle. Further expedient parameters are wind parameters, such as windvane offset for determination of inaccuracies or vortices in the area of a windvane of the wind energy installation. It is also possible to optimize other parameters, such as control parameters for control devices for the wind energy installation (for example PID regulators) or parameters for a control system for a converter for the wind energy installation (for example variance of the power, grid-system flicker or harmonics) using the method according to the invention. It has been proven to use not only one but a plurality of target variables. If, by way of example, one target variable to be optimized is the electrical power output, then it is also possible additionally to use as a target variable a measure of the structural load, for example on the rotor blades, and of the noise emission, and possibly to additionally provide a parameter for the machine load (damage equivalence load). The invention also relates to a wind energy installation for carrying out the method according to the invention. The wind energy installation has a pod which is arranged on a substructure that has a rotor which is arranged on its end face such that it can rotate, having a generator which is driven by the rotor in order to output electrical energy via a converter, with a control device and a measurement device being provided in order to carry out the method, having a microprocessor and a memory apparatus. The invention will be explained in more detail with reference to one exemplary embodiment, which is illustrated in the drawing, in which: Figure 1 shows an overall view of a wind energy installation for carrying out the method according to the invention; Figure 2 shows a partial view of the rotor of the wind energy installation illustrated in Figure 1; Figure 3 shows a schematic block diagram of the wind energy installation illustrated in Figure 1; Figure 4 shows an outline method for the optimization process according to the invention; Figure 5 shows a flowchart according to one exemplary embodiment of the optimization method according to the inventions-Figure 6 shows a detailed view of a variation of the method illustrated in Figure 5; and Figure 7 shows a detailed view of a further variation of the method illustrated in Figure 5. A wind energy installation which is intended to carry out the method according to the invention has a substructure which is in the form of a tower 1 and has a machine house 2 fitted to it. Most of the major components of the wind energy installation are arranged in or on this substructure. A rotor 3 is arranged on one end face of the machine house 2, such that it can rotate by means of a rotor shaft 50. Via this rotor shaft 50, the rotor 3 drives a generator 5 arranged in the machine house 2, possibly via a gearbox. The generator 5 may be in various forms, and may be a DC generator, a single-phase generator or a three-phase generator. Furthermore, it may be a synchronous machine or an asynchronous machine, with a single or double feed. The generator 5 is preferably in the form of a double-fed asynchronous machine. The generator 5 is connected to an electrical supply grid system 9 via a converter 7 and connecting lines which are not illustrated in any more detail. Furthermore, a control device 6 is provided, and controls the operation of the wind energy installation. A measurement device 23 is provided in order to determine the electrical power output. The measured electrical power is preferably the real power P, since this is a better measure than the volt-amperes or the wattless component of the power actually produced by the wind energy installation. However, it is also possible to provide for the electrical energy output to be used rather than the power. This has the advantage that integration is carried out automatically in the case of measurements over a relatively long time period. The meter that is provided in any case for the energy output in the wind energy installation (kilowatt-hours meter) can expediently be used for this purpose. This not only saves a separate sensor but automatically results in integration of the power signal, thus reducing undesirable noise in the measurement signal. It is also possible to use other parameters which represent a measure of the energy or power output from the rotor/generator system. In particular, it is possible to use the mechanical torque emitted from the rotor or the electrical torque produced by the generator, which can be used for formation of the efficiency measure, with or without rota- tion-speed assessment, depending on the operating range. The design of the control device 6 is illustrated schematically in Figure 3. The measurement device 23 for the electrical power output is connected to an analog/digital converter 61. This is designed to convert the signals supplied from the measurement device 23 to data in a data record at regular, variable time intervals. The data record is transmitted via a connecting line to a central control unit 60. The central control unit 60 has a microprocessor 62 for running the control and operating programs that are required for operation of the wind energy installation. A memory 64 is also connected to the central control unit 60. The generator 5 and the converter 7 are connected to outputs of the central control unit 60. Furthermore, a blade angle adjustment device 4 is connected to one output of the central control unit 60, and this will be explained in more detail in the following text. The blade angle adjustment device 4 is used to adjust the pitch angle a of the rotor blades 31. As is illustrated in Figure 2, the angle a is the angle included between a representative profile chord 39 of the rotor blade 31 and the rotation plane of the rotor, the normal to which is defined by the rotor shaft 50. Figure 2 shows a view of the blade tip 32 of one of the rotor blades 31 illustrated in Figure 3. The profile chord 39 connects the leading edge 35, which is located in the front area, to the trailing edge 34 of the rotor blade 31. The rotor blade 31 is anchored in the hub 4 by its blade root 33. The rotor blade 31 is twisted, such that the profile chord results in a pitch angle which be- comes continuously greater starting from the blade tip 32 to the blade root 33. The blade root 33 preferably has a circular shape in the area of the hub 4, so that the rotor blade 31 can be pivoted about the center point of the circle. The blade adjustment device 4 has a drive motor 40, which is arranged in the tip of the hub 4 and on whose power-output shaft an angle gearbox is arranged, comprising a large gearwheel 41 and a pinion 42. The pinion 42 engages in a circumferential tooth system which interacts with the blade root 33. This means that the rotor blade 31 can be pivoted on operation of the drive motor 40, as is symbolized by the double-headed arrow 4 9 illustrated in Figure 2. The procedure for the method according to the invention will now be explained, with reference to Figure 4a. The wind energy installation is initialized at the start 100 of the method. This means that interval limit values are set, old data is deleted from memories and, apart from this, the wind energy installation is prepared for operation in a manner known per se. Initial steps for the invention during the initialization process 102 are the starting of a timer t and the definition of an operating value ocB for a parameter to be optimized as a first value, with which operation will be started. This value can be defined in a manner known per se, as would be done in a wind energy installation without using the method according to the invention. After initialization, the optimization process 104 according to the invention can commence. Reference is now made to Figure 5. Starting from the operating value ocB an upper and a lower interval limit value aD and aL/ respectively, are each set in a first step 110, by adding an initial stress value OC0FF to the operating value aB, or subtracting it from the operating value αB- A sample counter i is set to zero. One of the two interval limit values is then set at the start of a cycle, the interval limit value αl in the illustrated exemplary embodiment (step 112). The rotor blades 31 are set to a first value α(i) = αl by means of the adjustment device 4. The wind energy installation is then operated using this pitch angle (step 114). In this case, the control device 6 uses the measurement device 23 to record the electrical energy output over a specific time period, possibly as well as other parameters- The measurement vector Z (αl) determined in this way is stored, after A/D conversion (if necessary) as a data record DL(i) in the memory 64 (step 116). In the next step 118, the pitch angle α(i) is changed to the upper interval limit value αd, and the wind energy installation is operated with this changed pitch angle (step 120). As before, a measurement vector is formed from data from the measurement device 23 and possibly from other parameters, and is stored in the memory 64, to be precise as a data record for the upper value Da(i) (step 122). This completes a sampling process with alternate operation of the wind energy installation using the upper and the lower interval limit value. In order to complete the process, the sample counter i is now incremented by one step (step 124). If the number of samples taken in this way is less than a predetermined limit n, then samples are once again taken from the step 112 until the number of samples is equal to the preselectable value n (step 126). By way of example, the values of n are in the range between 100 and 100 000, and a range between 1 000 and 10 000 has been particularly proven. The number is preferably variable, for example as a function of the variance of the parameters included in the data records. The time period over which a sample is taken is expediently chosen as a function of the parameter under consideration. In the illustrated exemplary embodiment, using the pitch angle a, the time duration is expediently between 10 and 120 seconds. If a torque characteristic is used as the parameter, the time period is expediently between 30 seconds and 10 minutes, and for control parameters it is expediently between 10 seconds and 10 minutes. For more dynamic variables, such as converter parameters, a shorter time period is preferable, for example of between 10 milliseconds and 30 seconds. The choice of this time period is based on the knowledge that it is preferably chosen until non-stationary effect are eliminated, such as those caused by inertia of the air flow (dynamic wake). After completion of the sampling process, the evaluation process starts with step 128. The data records contained in the memory 64 for the lower and upper pitch angles DL(i) and D0(i) are called up and evaluated for all (i) from 0 to n-1. The evaluation process may relate to all of the elements in the data record, or else only to some of them. In the first-mentioned case, the individual elements are expediently linked by means of weighting coefficients to form a scalar variable. In the second-mentioned case, the calculation is carried out using the magnitude directly. The quality measure may be a complex, multiple-parameter, non-linear function. However, a simple quality measure may also be provided, for whose formation there is no need to use all of the stored individual values. A quality measure such as this has the advantage that it can be formed continuously while the sampling cycle is being carried out. The measured values are in this case processed directly in order to form the quality measure. This has the advantage that there is no need to store the measured values. One example of a quality measure such as this is a value sum, for example for the energy fed into the grid system from a wind energy installation during one cycle, with the wind energy installation being operated in the region of the optimum tip-speed ratio. The energy values are integrated separately on the basis of the interval limit values; this can be done continuously during the measurement, so that there is no need for storage. In order to determine whether the wind energy installation is actually being operated in the region of the optimum tip-speed ratio while the samples are being taken, an operating-mode detector 22 is expediently provided. This is designed to use the rotation speed of the rotor 3 to determine whether the installation is being operated in this range. The detector 22 is expediently in the form of a threshold-value switch, having a lower rotation speed limit and an upper rotation speed limit, between which the operating range of optimum tip-speed ratio is located. The detector 22 does not necessarily need to use the rotation speed as an input signal, and the electrical power output from the wind energy installation can also be used for determination purposes. However, an appropriate status signal for the operating mode is also frequently already available in the machine control system 6. It is then sufficient to read this signal. There is then no need for any separate sensors. The detector 22 ensures that the only measured values which are used as samples for the statistical processes are those in which the wind energy installation has been operated in the region of the optimum tip-speed ratio. If this is not the case, then the sample is invalid, and cannot be used. It is rejected and the sampling process is repeated. Further simplifications of the method according to the invention are possible. For example, the measurement device 23, the A/D converter 61 and the memory 64 do not necessarily need to be provided- The measure for the electromechanical quantity and, possibly, the other parameters that are used are frequently available in any case from the operating control system for the wind energy installation, to be precise in many cases already in the form of mean values, as well. There is then no need for any separate measurement and storage. The parameters can be further-processed directly using the method according to the invention. At the end of step 128, values GL and GD are produced as quality measures for the two interval limit values. A check is carried out in a next step 130 to determine whether the two quality measures Gv and GL are of equal magnitude (with a tolerance e), or not- If they are not equal, a check is carried out to determine which of the two interval limit values has the better associated quality measure. If the upper interval limit value has achieved the better quality measure, the operating value αlis incremented by the step width A (step 134). If the lower interval limit value has achieved the better quality measure, then the operating value αb is decremented by the step width A (step 136). After a backward jump 137, the process starts once again with the step 110, and is continued until an optimum value is finally achieved. This is determined by comparison of the two quality measures in step 132- In the event of a positive result, the optimization process has been successful, and a changeover is made to control operation 138. The optimization process thus reaches its end 140. The rest of the procedure is then as illustrated in Figure 4a. Returning now to Figure 4a again, the wind energy installation is operated using the optimum value determined in this way (step 108). This is continued (return path 109) until a variable repetition time tw has elapsed (step 106). This is followed by a backward jump 107 in order to carry out the optimization process (104) again. The wait for a predeterminable repetition time to have elapsed in step 106 does not need to be the only criterion for repetition of the optimization process 104 according to the invention. Further criteria can alternatively or additionally be provided (see Figure 4b). For example, Figure 4b shows an extended form (step 1061) according to which an investigation is additionally carried out into significant changes in one of the parameters used, for example the electrical power. If a limit value ]i which can be preset is exceeded, then a backward jump 107 is also made even if the repetition time has not yet elapsed. The invention is not restricted to the exemplary embodiment illustrated in Figure 5. Various variations of the optimization process according to the invention are feasible. By way of example, Figure 6 illustrates two mutually independent variations. The first relates to variation of the interval limit values during the iteration. In contrast to the situation in step 134, the operating value αb and thus the two interval limit val- ues are not shifted in step 1341, but new interval limit values are fixed directly, and are each incremented by the step value A. The second variation relates to a change in the interval width. This can be done, for example in the manner illustrated in step 134'. The lower interval limit value aL is increased by the step value A without being changed. The upper interval limit value aD is, however, not increased by the same amount, but the magnitude is reduced by a reduction value 5. In this case, 5 may also have the same value as A, that is to say the upper interval limit value would remain unchanged. The interval defined by the two start values is therefore reduced by the reduction value 5 so that the method according to the invention converges more rapidly. In a corresponding manner, in step 135', the upper interval limit value is reduced by the step value A without being changed; while the lower interval limit value αl is reduced only by A-6. It is self-evident that, in this variant, the interval limit values are not initialized once again in step 110. In a corresponding manner, it is alternatively also possible to provide for the scatter value a0FF to be reduced by half the reduction value 6/2. This likewise results in a reduction in the interval by the reduction value 5. A predictor is preferably used in order to define the step values A and, if appropriate, 5. As is illustrated in Figure 7, two predictor functions θ and θ are used for this purpose in step 131 on the basis of the determined values for the quality measure, as well as the interval limit values, if appropriate also including the data records contained in the memory 64. The step value A and the reduction value 6 can be matched to the respective circumstances with the aid of the predictor functions. It is also possible to provide for the step value A and the reduction value 5 to be determined on the basis of characteristics or of look-up tables, instead of by the use of a predictor. This has the advantage that empirical values relating to the choice of the step values A and the reduction values 5 can be taken into account at the start and end of the optimization process according to the invention. Furthermore, this offers the advantage of simple calculation. If appropriate, it is thus possible to use higher step values at the start of the method, in order to achieve faster convergence, while smaller step values are used later on, in order to achieve greater accuracy. Patent Claims 1. A method for optimization of operating parameters of a wind energy installation, in particular with respect to its rotor/generator system, with definition of an upper and a lower interval limit value (110) for a parameter to be optimized, characterized by a cycle with alternate operation (112, 114, 118, 120) of the wind energy installation with the interval limit values, with one data record in each case being produced with a target variable (116, 122), to be precise over a variable number of repetitions (126), evaluation of the data records relating to the interval limit values with a quality measure being formed, identification of the interval limit value with the poorer quality measure (130), replacement of at least this interval limit value by shifting through a step value A in the direction of the other interval limit value (134, 136), repetition of the cycle (137). 2. The method as claimed in claim 1, characterized in that the interval limit values are predetermined directly. 3. The method as claimed in claim 1, characterized in that the interval limit values are defined by an operating value (aB) and a scatter value (CXOFF) • 4. The method as claimed in claim 3, characterized in that the operating value (aB) is varied through the step width A in order to vary at least the poorer interval limit value. 5. The method as claimed in one of claims 1 to 4, characterized in that the separation between the interval limit values is reduced by a reduction value 5, preferably by reducing the scatter value αoff 6. The method as claimed in one of claims 1 to 5, characterized in that the cycle is repeated until a predeterminable termination criterion (E) is reached (130) . 7. The method as claimed in claim 5, characterized in that the cycle is repeated until the interval limit values have reached a separation of (e1). 8. The method as claimed in claim 3 or 4, characterized in that a predictor is used to determine the step value A and, possibly, the reduction value 8 (131). The method as claimed in one of the preceding claims, characterized in that the statistical quality measure is calculated by addition (128). The method as claimed in one of the preceding claims, characterized in that the data record contains a plurality of target variables, and the quality measure is formed multidimensionally. The method as claimed in one of the preceding claims, characterized in that the parameter comprising the pitch angle of the rotor blade and the target variable is the electrical power or energy output. The method as claimed in one of claims 1 to 11, characterized in that the parameter is a torque characteristic. The method as claimed in one of claims 10 to 12, characterized in that the structural load and/or the noise emission are/is additionally used as target variable or variables. Wind energy installations having a machine house (2), which is arranged on a substructure (1), and having a rotor (3) which is arranged on the end face of the machine house (2) such that it can rotate, having a generator (5) which is driven by the rotor (3) in order to output electrical energy via a converter (7), and having a control device (6), which has a microprocessor (60) and a memory device (64), as well as a measurement device (23), characterized in that the control device (6) is designed to carry out the method as claimed in one of claims 1 to 13-

Documents:

2700-CHENP-2007 CORRESPONDENCE OTHERS 16-05-2011.pdf

2700-CHENP-2007 POWER OF ATTORNEY 15-02-2012.pdf

2700-CHENP-2007 AMENDED PAGES OF SPECIFICATION 15-02-2012.pdf

2700-CHENP-2007 AMENDED CLAIMS 15-02-2012.pdf

2700-CHENP-2007 CORRESPONDENCE OTHERS.pdf

2700-CHENP-2007 CORRESPONDENCE PO.pdf

2700-CHENP-2007 EXAMINATION REPORT REPLY RECEIVED 15-02-2012.pdf

2700-CHENP-2007 FORM-1 15-02-2012.pdf

2700-CHENP-2007 FORM-18.pdf

2700-CHENP-2007 FORM-3 15-02-2012.pdf

2700-chenp-2007-abstract.pdf

2700-chenp-2007-claims.pdf

2700-chenp-2007-correspondnece-others.pdf

2700-chenp-2007-description(complete).pdf

2700-chenp-2007-drawings.pdf

2700-chenp-2007-form 1.pdf

2700-chenp-2007-form 3.pdf

2700-chenp-2007-form 5.pdf

2700-chenp-2007-pct.pdf