Model Identification Method for Hydropower Station Turbine Using Computer-Generated Random Excitation

Abstract

The present invention provides a method for identifying a turbine model of a hydropower station using computer-generated random excitation, which first designs a PID controller so that the frequency of the entire turbine can stably track the expected frequency, and then performs parameter identification under closed-loop stability. First, the turbine frequency, relay stroke, etc. are measured by sensor, and then the expected frequency data that incorporates the computer-generated random excitation source is set to continuously excite the entire system. Then, an identification model of the relay and generator set to be identified is established. Then, the output of the identification model is compared with the output of the real model to obtain identification error data. Then, based on the identification error data and the identification model structure, a parameter identification law design method of a multi-parameter hinge coupling is designed, so that the identification process can quickly break the non-random balance and can effectively improve the identification speed and identification effect. This method has the advantages of simple implementation, high identification accuracy, and a wide range of identification parameters.

Description

Technical Field

[0001] The present invention relates to the field of model parameter identification, and in particular to a method for identifying model parameters of a water turbine in a hydropower station. Background Art

[0002] Hydroelectric turbine generators are a crucial component of hydropower generation at hydropower stations. Besides ensuring safe power supply, they also require voltage and frequency to be stabilized within a narrow range of given values. The turbine control system, structurally comprised of components such as the water diversion system, servo system, speed governor, turbine, and generator, encompasses complex nonlinear control systems, including hydroelectric machinery. Due to objective constraints, accurate modeling and real-machine testing are difficult. Therefore, analyzing and studying model parameter identification methods and conducting identification tests to establish accurate models for different scenarios can lay the foundation for stable unit operation and have important engineering value and practical significance. PID control is currently widely used in turbine control and regulation systems. Identifying more accurate system models and parameters would provide a better theoretical basis for turbine control and PID parameter tuning, thereby improving the dynamic performance of hydropower units and achieving more satisfactory unit control results. Currently, methods used for parameter identification include gravitational search algorithms, ant lion optimization algorithms, and neural network algorithms. However, these algorithms are often complex, require high data volumes, and are computationally intensive. In the above background, the present invention proposes a turbine model parameter identification method based on the combination of random theory and adaptive parameter identification, which has the advantages of simple and convenient operation, small calculation amount and high identification accuracy.

[0003] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present invention, and therefore may include information that does not constitute prior art known to ordinary technicians in this field. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for identifying a turbine model of a hydropower station using computer-generated random excitation, thereby overcoming the problems of low accuracy and limited applicability of traditional model parameter identification to a certain extent.

[0005] According to one aspect of the present invention, a method for identifying a model of a hydropower station turbine using computer-generated random excitation is provided, comprising the following steps:

[0006] Step S10: Install an EPC-775A photoelectric encoder on the rotating shaft of the turbine generator set to measure the frequency of the turbine generator set, install a sensor at the output end of the relay to measure the relay stroke, install a sensor at the output end of the turbine to measure the turbine output power, set a unified discrete period, and convert the above data into unified period discrete data; respectively obtain the frequency discrete data of the turbine generator set, the relay stroke discrete data, and the turbine output power discrete data.

[0007] Step S20, setting the expected frequency of the turbine, then using a computer to generate a random excitation source frequency, superimposing the two to obtain the identification expected frequency of the turbine; and then converting it into discrete data of the identification expected frequency of the turbine according to a unified discrete period.

[0008] In step S30, the expected frequency discrete data of the turbine is compared with the frequency discrete data of the turbine generator set to obtain the frequency error discrete data of the turbine generator set; then, a mixed operation of proportion, integration, and difference is performed on the error data to form a turbine PID control to obtain the control discrete data of the turbine generator set.

[0009] In step S40, a servomotor parameter identification model is established and the output data of the servomotor identification model is calculated. The output data is then compared with the discrete data of the servomotor stroke to obtain servomotor identification error data. Based on the servomotor identification error data, control-related parameter identification laws and servomotor-related parameter identification laws are designed, respectively, and the identification results of the servomotor parameters are integrated. The servomotor amplification gain parameter and the servomotor time constant are then inversely solved according to the formula.

[0010] Step S50: Establish a generator set parameter identification model and solve the output data of the generator set identification model; then compare it with the frequency discrete data of the turbine generator set to obtain the generator set identification error data; then, based on the generator set identification error data, respectively design the generator set input-related parameter identification law and the generator set output-related parameter identification law and integrate them to obtain the generator set parameter identification result; then, inversely solve the generator set amplification gain parameter and the generator set inertia time constant according to the formula.

[0011] In an exemplary embodiment of the present invention, a desired frequency of the turbine is set, and then a random excitation source frequency is generated by a computer. The two are superimposed to obtain the identification desired frequency of the turbine. The identification desired frequency of the turbine is then converted into discrete data according to a uniform discrete period. The data includes:

[0012] f r =f r0 +f rr (R-0.5);

[0013] where f r0is the expected frequency of the turbine, R is a random number between the interval [0,1] generated by a computer; f rr is the amplitude of the random excitation source frequency generated by computer; f r is the expected identification frequency of the turbine; f r (n) is the expected frequency discrete data of the turbine identification.

[0014] In an exemplary embodiment of the present invention, the identified expected frequency discrete data of the turbine is compared with the frequency discrete data of the turbine generator set to obtain the frequency error discrete data of the turbine generator set; then, a proportional, integral, and differential mixed operation is performed on the error data to form a turbine PID control including:

[0015] e(n)=f r (n)-f(n);

[0016]

[0017]

[0018] u(n)=k p e(n)+k s s(n)+k d D(n);

[0019] Where f(n) is the frequency discrete data of the turbine generator set, e(n) is the frequency error discrete data of the turbine generator set; s(n) is the integral discrete data of the frequency error of the turbine generator set; D(n) is the differential discrete data of the frequency error of the turbine generator set; u(n) is the PID control discrete data of the turbine generator set; k p 、k s 、k d are the control parameters of error proportion, integration and difference.

[0020] In an exemplary embodiment of the present invention, a servomotor parameter identification model is established, and output data of the servomotor identification model is calculated. This data is then compared with discrete servomotor stroke data to obtain servomotor identification error data. Based on the servomotor identification error data, control-related parameter identification laws and servomotor-related parameter identification laws are designed, respectively, and the identification results of the servomotor parameters are integrated. The servomotor amplification gain parameter and the servomotor time constant are then inversely solved according to the formula:

[0021] y bn (n) = c 6g (n)u(n)+c 7g (n)y(n-1);

[0022] e j (n) = y(n) - y bn(n);

[0023]

[0024]

[0025]

[0026]

[0027]

[0028]

[0029] where y bn (n) is the output data of the relay identification model, e j (n) is the relay identification error data; y(n) is the relay stroke discrete data, c 6d (n) is the identification law of control related parameters; k c6 、k c6a , ε1 is the speed adjustment coefficient for control related parameter identification; c 7d (n) is the identification law of the servomotor related parameters; k c7 、k c7a , ε2 is the speed adjustment coefficient for identifying the relevant parameters of the relay; c 6g (n) is the control related parameter identification data; c 7g (n) is the identification result data of the relay related parameters; T g (n) is the relay time constant; k a (n) is the relay amplification gain parameter.

[0030] In an exemplary embodiment of the present invention, a generator set parameter identification model is established, and output data of the generator set identification model is calculated; then, the output data is compared with the frequency discrete data of the hydro-turbine generator set to obtain generator set identification error data; then, based on the generator set identification error data, identification laws for the generator set input-related parameters and the generator set output-related parameters are designed and integrated to obtain the generator set parameter identification results; then, the generator set amplification gain parameter and the generator set inertia time constant are inversely solved according to the formula, including:

[0031] f bn (n) = c 1g (n)P(n)+c 2g (n)f(n-1);

[0032] e f (n) = f(n) - f bn (n);

[0033]

[0034]

[0035]

[0036]

[0037]

[0038]

[0039] Where P(n) is the discrete data of turbine output power, f bn (n) is the output data of the generator set identification model, e f (n) is the generator set identification error data; c 1d (n) is the identification law of the relevant parameters of the generator set input; k c1 、k c1a , ε3 is the input related parameter identification speed adjustment coefficient; c 2d (n) is the identification law of the generator set output related parameters; k c2 、k c2a , ε4 is the speed adjustment coefficient for identifying the relevant parameters of the generator set output; c 1g (n) inputting relevant parameter identification data for the generator set; c. 2g (n) is the identification result data of the relevant parameters of the generator set output; T s (n) is the inertia time constant of the generator set; k s (n) is the generator set amplification gain parameter.

[0040] Beneficial effects

[0041] The present invention provides a method for model parameter identification of a hydropower station turbine using computer-generated random excitation. One of the effects is that it can greatly reduce the problem of low control quality of hydropower station turbines caused by model uncertainty. The second effect is that the method of model parameter identification is simpler and more convenient than the physical modeling process, which can greatly save time and has higher economic benefits. The method has the following innovations. First, it proposes a method of using computer-generated random excitation to greatly improve the accuracy of traditional parameter identification, as well as the scope and breadth of application of parameter identification. We have compared it in experiments and found that when random excitation is not used, some parameters cannot be accurately identified. The use of random excitation makes good use of the variability of the random process, thereby utilizing the statistical characteristics of the random process, which can greatly improve the effect of parameter identification. The second innovation is to propose a parameter identification law design method for multi-parameter related hinges, so that the identification process can quickly break the non-random equilibrium, which can effectively improve the identification speed and identification effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The accompanying drawings are incorporated into and constitute a part of this specification, illustrate embodiments consistent with the present invention, and together with the description, serve to explain the principles of the present invention. Obviously, the drawings described below are only some embodiments of the present invention, and it is clear that those skilled in the art can derive other drawings based on these drawings without inventive effort.

[0043] Figure 1 This is a flow chart of a method for identifying a hydropower station turbine model using computer-generated random excitation provided by the present invention;

[0044] Figure 2 Schematic diagram of the hydropower station turbine model identification structure of the method provided by the embodiment of the present invention;

[0045] Figure 3 is a discrete data curve of frequency error of a hydro-turbine generator set according to the method provided in an embodiment of the present invention;

[0046] Figure 4 is a servomotor identification model error curve of the method provided in the embodiment of the present invention;

[0047] Figure 5 is a curve of identification results of the relay amplification gain parameters of the method provided in the embodiment of the present invention;

[0048] Figure 6 is a curve of the identification result of the relay time constant of the method provided in the embodiment of the present invention;

[0049] Figure 7 is an identification curve of the relay amplification gain parameter when the method provided in the embodiment of the present invention does not adopt random excitation;

[0050] Figure 8 is an identification curve of the servomotor time constant when the method provided in the embodiment of the present invention does not adopt random excitation;

[0051] Figure 9 is a frequency curve of a hydro-turbine generator set according to a method provided by an embodiment of the present invention;

[0052] Figure 10 is a generator set model identification error curve of the method provided by the embodiment of the present invention;

[0053] Figure 11 is a generator set amplification gain parameter identification curve of the method provided in an embodiment of the present invention;

[0054] Figure 12 This is an enlarged diagram of the terminal of the generator set amplification gain parameter identification result of the method provided in the embodiment of the present invention;

[0055] Figure 13 is the generator set inertia time constant identification curve of the method provided in the embodiment of the present invention;

[0056] Figure 14 This is an enlarged view of the terminal of the generator set inertia time constant identification result of the method provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0057] The present invention provides a method for identifying a turbine model of a hydropower station using computer-generated random excitation. The method first designs a PID controller so that the frequency of the entire turbine can stably track the desired frequency, and then performs parameter identification under closed-loop stability. First, the desired frequency data that incorporates a computer-generated random excitation source is set to perform PID feedback control and continuous excitation on the entire system. Then, an identification model of the relay and generator set to be identified is established. Then, the output of the identification model is compared with the output of the real model to obtain identification error data. Then, a parameter identification law design method of a multi-parameter hinge coupling is designed, so that the identification process can quickly break the non-random balance and effectively improve the identification speed and identification effect. The method has the advantages of simple implementation, high identification accuracy, and a wide range of identification parameters.

[0058] The following will further explain and illustrate the hydropower station turbine model identification method using computer-generated random excitation of the present invention with reference to the accompanying drawings. Figure 1 and Figure 2 As shown, the hydropower station turbine model identification method using computer-generated random excitation includes the following steps:

[0059] Step S10 is specifically divided into the following four steps: First, an EPC-775A photoelectric encoder is installed on the rotating shaft of the hydro-turbine generator set to measure the frequency of the hydro-turbine generator set, which is recorded as f.

[0060] The second step is to install a sensor at the output end of the relay to measure the relay stroke, which is recorded as y.

[0061] The third step is to install a sensor at the output end of the turbine to measure the turbine output power, which is recorded as P.

[0062] The fourth step is to set the uniform discrete period to T0, convert the above data into uniform period discrete data, and obtain the frequency discrete data of the turbine generator set, the relay stroke discrete data, and the turbine output power discrete data respectively; denoted as f(n), y(n), and P(n), which represent the values ​​at time t=n*T0 respectively.

[0063] Step S20 is specifically divided into the following three steps: The first step is to set the desired frequency of the turbine, which is less than 1 / 3 of the normal operating frequency.

[0064] In the second step, a computer is used to generate a random excitation source frequency, and the two are superimposed to obtain the expected identification frequency of the turbine as follows:

[0065] f r =f r0 +f rr (R-0.5);

[0066] where f r0 is the expected frequency of the turbine, R is a random number between the interval [0,1] generated by a computer; f rr is the amplitude of the random excitation source frequency generated by computer; f r is the expected frequency for turbine identification.

[0067] The third step is to convert it into the identification expected frequency discrete data of the turbine according to the unified discrete period to obtain f r (n) is the expected frequency discrete data of the turbine identification.

[0068] Step S30 can be specifically divided into the following four steps: The first step is to compare the expected frequency discrete data of the turbine with the frequency discrete data of the turbine generator set to obtain the frequency error discrete data of the turbine generator set as follows:

[0069] e(n)=f r (n)-f(n);

[0070] Where e(n) is the discrete data of the frequency error of the turbine generator set.

[0071] In the second step, the integral operation is performed based on the discrete data of the generator set frequency error to obtain the integral discrete data of the turbine generator set frequency error as follows:

[0072]

[0073] Where s(n) is the discrete data of the frequency error integral of the turbine generator set.

[0074] In the third step, a differential operation is performed based on the discrete data of the generator set frequency error to obtain the differential discrete data of the turbine generator set frequency error as follows:

[0075]

[0076] Where D(n) is the frequency error differential discrete data of the turbine generator set.

[0077] In the fourth step, the three above are linearly superimposed to obtain the discrete control data of the turbine generator set; this data is transmitted to the input end of the turbine servo, and the control parameters of the error proportion, integration, and difference are adjusted to enable the turbine generator set to operate normally, providing a stable identification environment for the turbine parameter identification as follows:

[0078] u(n)=k p e(n)+k s s(n)+k d D(n);

[0079] Where u(n) is the discrete data of turbine generator control; k p 、k s 、k d are the control parameters of error proportion, integration and difference.

[0080] It is important to note that direct open-loop parameter identification can also be performed in an uncontrolled state; however, its applicability, stability, and robustness are not as good as those achieved under closed-loop PID control. Therefore, the present invention performs parameter identification based on closed-loop control. In particular, when random excitation is introduced by the computer, the open-loop identification effect is even worse than the closed-loop identification effect.

[0081] Step S40 can be specifically divided into the following four steps: The first step is to establish a servomotor parameter identification model based on the turbine generator control discrete data and the servomotor stroke discrete data, and calculate the output data of the servomotor identification model as follows:

[0082] y bn (n) = c 6g (n)u(n)+c 7g (n)y(n-1);

[0083] where y bn (n) is the output data of the relay identification model.

[0084] In the second step, the relay identification error data is compared with the discrete data of the relay stroke, and the following is obtained:

[0085] e j (n) = y(n) - y bn (n);

[0086] where e j (n) is the relay identification error data.

[0087] In the third step, the control-related parameter identification law and the relay-related parameter identification law are designed according to the relay identification error data as follows:

[0088]

[0089]

[0090] where c 6d (n) is the identification law of control related parameters; k c6 、k c6a , ε1 is the speed adjustment coefficient for control related parameter identification; c 7d (n) is the identification law of the servomotor related parameters; k c7 、k c7a , ε2 is the speed adjustment coefficient for identifying the relevant parameters of the relay.

[0091] In the fourth step, integration is performed to obtain the identification result data of the control-related parameters and the identification result data of the relay-related parameters as follows:

[0092]

[0093]

[0094] where c 6g (n) is the control related parameter identification data; c 7g (n) is the identification result data of the relay related parameters.

[0095] Step 5: According to the formula, the relay amplification gain parameter and relay time constant are solved as follows:

[0096]

[0097]

[0098] Where T g (n) is the relay time constant; k a (n) is the relay amplification gain parameter; where k of the relevant hinge c6a 、k c7a The design of coefficient parameter identification law enables the identification process to quickly break the non-random balance, thereby improving the identification speed and identification effect.

[0099] Step S50 can be specifically divided into the following five steps: The first step is to establish a generator set parameter identification model based on the frequency discrete data and turbine output power discrete data of the turbine generator set, and calculate the output data of the generator set identification model as follows:

[0100] f bn (n) = c 1g (n)P(n)+c 2g (n)f(n-1);

[0101] where f bn(n) is the output data of the generator set identification model.

[0102] The second step is to compare the discrete frequency data of the turbine generator set and obtain the generator set identification error data as follows:

[0103] e f (n) = f(n) - f bn (n);

[0104] where e f (n) is the generator set identification error data.

[0105] The third step is to design the identification laws of the generator set input-related parameters and the generator set output-related parameters according to the generator set identification error data as follows:

[0106]

[0107]

[0108] where c 1d (n) is the identification law of the relevant parameters of the generator set input; k c1 、k c1a , ε3 is the input related parameter identification speed adjustment coefficient; c 2d (n) is the identification law of the generator set output related parameters; k c2 、k c2a , ε4 is the speed adjustment coefficient for identifying the relevant parameters of the generator set output.

[0109] In the fourth step, the identification result data of the generator set input related parameters and the identification result data of the generator set output related parameters are obtained by integration as follows:

[0110]

[0111]

[0112] where c 1g (n) inputting relevant parameter identification data for the generator set; c. 2g (n) is the identification result data of the generator set output related parameters.

[0113] Step 5: According to the formula, the amplification gain parameter of the generator set and the inertia time constant of the generator set are inversely solved as follows:

[0114]

[0115]

[0116] Where T s (n) is the inertia time constant of the generator set; ks (n) is the generator set amplification gain parameter.

[0117] At this point, the identification of relevant parameters of the turbine model is completed. The following two cases are provided to illustrate the specific implementation process and results.

[0118] Case Implementation 1

[0119] To verify the feasibility and effectiveness of the above method, the following case simulation analysis was conducted. To reduce mutual interference, the servomotor parameter identification and the turbine generator set parameter identification were performed separately. In case study one, the servomotor parameter identification was performed. In case study two, the turbine generator set parameter identification was performed.

[0120] In step S10 , the uniform discrete period is set to T0 = 0.005.

[0121] In step S20, the present invention selects f r0 =5; in the present invention, f is selected rr =2.5.

[0122] In step S30, in the present invention, k is selected p =2, k s =0.3, k d =2.5; get the discrete data of frequency error of turbine generator set as Figure 3 shown.

[0123] In step S40, in the present invention, k is selected c6 =2.5, k c6a =0.2, ε1=0.3, k c7 =2.5, k c7a =0.2, ε2=0.3. Finally, the error curve of the relay identification model is obtained as follows: Figure 4 As shown in the figure, the identification results of the relay amplification gain parameters are as follows: Figure 5 As shown, its standard value is 1.05, and the identification result of the relay time constant is as follows Figure 6 As shown in the figure, the standard value is 28.5. It can be seen that in the case of computer random excitation, although it takes a longer time, the accuracy of parameter identification is higher. When random excitation is not used, the identification results of the relay amplification gain parameters are as follows: Figure 7 As shown in the figure, the identification results of the relay time constant are as follows: Figure 8 It can be seen that the identification result of the relay time constant is very poor, while the identification result of the relay amplification gain parameter is acceptable.

[0124] Case Implementation 2

[0125] In step S10 , the uniform discrete period is set to T0 = 0.01.

[0126] In step S20, the present invention selects f r0 =6; in the present invention, f is selected rr =2, the frequency of the turbine generator set is obtained as Figure 9 shown.

[0127] In step S50, the present invention selects k c1 =0.5, k c1a =0.2, ε1=0.3, k c7 =0.5, k c7a =0.2, ε2=0.3. The generator model identification error is obtained as Figure 10 shown.

[0128] At the same time, the identification results of the generator set amplification gain parameters are obtained as follows: Figure 11 As shown, the terminal enlarged diagram is as follows Figure 12 As shown, its standard value is 20, and the identification result of the generator set inertia time constant is as follows Figure 13 As shown, the terminal enlarged diagram is as follows Figure 14 As shown in the figure, the standard value is 3.7. It can be seen that the accuracy of the final parameter identification result fluctuates around the standard value. As time goes by, it will tend to be stable, and the accuracy of parameter identification will become higher and higher.

[0129] The above implementation cases show that the turbine model parameter identification method provided by the present invention is completely feasible and can provide valuable model parameter experience data for further improving its control quality, thus having high engineering application value.

Claims

1. A hydropower station turbine model identification method using computer-generated random excitation is characterized by: The following steps are involved: Step S10: Install an EPC-775A photoelectric encoder on the shaft of the turbine generator set to measure the frequency of the turbine generator set, which is recorded as , install a sensor at the output end of the relay to measure the relay stroke, recorded as , install a sensor at the output end of the turbine to measure the turbine output power, which is recorded as ; Set a unified discrete period, convert the above data into a unified period discrete data, and obtain the frequency discrete data of the turbine generator set, the relay stroke discrete data, and the turbine output power discrete data, which are recorded as 、 、 ; Step S20: Set the expected frequency of the turbine, then use a computer to generate a random excitation source frequency, superimpose the two, and obtain the identification expected frequency of the turbine; then convert it into discrete data of the identification expected frequency of the turbine according to a unified discrete period as follows: ； in is the desired frequency of the turbine, A random number between the interval [0,1] generated by a computer; The amplitude of the random excitation source frequency is generated by computer; is the expected frequency of turbine identification; is the expected frequency discrete data of the turbine identification, representing time The value of is a unified discrete period; Step S30: Compare the identified expected frequency discrete data of the turbine with the frequency discrete data of the turbine generator set to obtain the frequency error discrete data of the turbine generator set; then perform a proportional, integral, and differential mixed operation on the error data to form a turbine PID control to obtain the control discrete data of the turbine generator set; Step S40: Establish a servomotor parameter identification model, calculate the output data of the servomotor identification model, and then compare the output data with the discrete data of the servomotor stroke to obtain servomotor identification error data. Then, based on the servomotor identification error data, design control-related parameter identification laws and servomotor-related parameter identification laws, and integrate the identification results of the servomotor parameters. Then, the relay amplification gain parameter and the relay time constant are inversely solved according to the formula; Step S50, establishing a generator set parameter identification model and calculating output data of the generator set identification model; Then, the frequency discrete data of the turbine generator set is compared to obtain the generator set identification error data; Then, according to the generator set identification error data, the generator set input related parameter identification law and the generator set output related parameter identification law are designed respectively and integrated to obtain the generator set parameter identification result; Then, the generator set amplification gain parameter and the generator set inertia time constant are inversely solved according to the formula.

2. The method for identifying a hydropower station turbine model using computer-generated random excitation according to claim 1, characterized in that: A servo parameter identification model is established, and the output data of the servo identification model is solved. This data is then compared with the discrete data of the servo stroke to obtain the servo identification error data. Based on the servo identification error data, the control-related parameter identification law and the servo-related parameter identification law are designed respectively, and the identification results of the servo parameters are integrated. The servo amplification gain parameter and the servo time constant are then inversely solved according to the formula: ； ； ； ； ； ； ； ； in Discrete data for turbine generator control, is the output data of the relay identification model, Identify error data for the relay; is the discrete data of relay stroke, To identify the laws of control related parameters; 、 、 Identify the speed adjustment coefficient for control related parameters; Identification law of servomotor related parameters; 、 、 Identify the speed adjustment coefficient for the servomotor's related parameters; Identify data for control related parameters; The identification result data of the relay-related parameters; is the servomotor time constant; is the relay amplification gain parameter.

3. The method for identifying a hydropower station turbine model using computer-generated random excitation according to claim 2, characterized in that: Establish a generator set parameter identification model and solve the output data of the generator set identification model; Then, the frequency discrete data of the turbine generator set is compared to obtain the generator set identification error data; Then, according to the generator set identification error data, the generator set input-related parameter identification law and the generator set output-related parameter identification law are designed respectively and integrated to obtain the generator set parameter identification results; then, according to the formula, the generator set amplification gain parameter and the generator set inertia time constant are inversely solved, including: ； ； ； ； ； ； ； ； in is the frequency discrete data of the turbine generator set, representing its Moment and The value of the moment, is the discrete data of turbine output power, is the output data of the generator identification model, Identify error data for the generator set; Input relevant parameter identification laws for the generator set; Identify the speed adjustment coefficient for input related parameters; Output relevant parameter identification law for the generator set; 、 、 Identify the speed adjustment coefficient for the generator set output related parameters; Input relevant parameter identification data for the generator set; Output identification result data of relevant parameters for the generator set; is the inertia time constant of the generator set; It is the generator set amplification gain parameter.

Images