Partition parameter identification method for network construction type doubly-fed wind turbine generator

Abstract

The invention relates to a partition parameter identification method for a network construction type doubly-fed wind turbine generator, solves the problem of how to accurately identify machine side system control parameters of the network construction type doubly-fed wind turbine generator, and belongs to the technical field of new energy power generation parameter identification. The method comprises the following steps: building a transient model of the network-building type double-fed fan according to a system structure of the network-building type double-fed fan, and obtaining a target error function of the transient model of the network-building type double-fed fan; the method comprises the following steps of: partitioning output data of the transient model of the network-forming doubly-fed fan about original machine side control parameters to obtain a plurality of intervals, increasing values of the original machine side control parameters, calculating distance correlation coefficients of all the machine side control parameters about each interval before and after the values are increased, and calculating the distance correlation coefficients of all the machine side control parameters; selecting the highest correlation interval as an identification interval of the to-be-identified parameter according to the distance correlation coefficient; and the target error function value is used as the fitness, a differential evolution algorithm is adopted to carry out optimization solution in the identification interval, and the optimization result is the machine side control parameter in the transient model of the grid-forming type doubly-fed fan.

Description

Technical Field

[0001] The invention relates to a method for identifying partition parameters of a grid-type doubly-fed wind turbine generator set, and belongs to the technical field of parameter identification of new energy power generation. Background Art

[0002] In recent years, with the large-scale access of wind power to the power grid, the wind power penetration rate of the power system has continued to increase. Due to the characteristics of weak damping and low inertia of the grid-following doubly-fed wind turbines, a large number of transient instability problems have occurred under weak power grids, posing a huge challenge to the safe and stable operation of the power grid. In order to improve the weak power grid operation and active support capabilities of doubly-fed wind turbines, grid-building doubly-fed wind turbines have been proposed and received widespread attention.

[0003] At present, power grid analysis mainly relies on digital simulation or hybrid simulation. Due to the manufacturer's confidentiality, it is difficult to obtain the correct control parameters, which makes it difficult to accurately analyze the fault characteristics. Accurate fault characteristics provide a theoretical basis for new protection principles and protection setting calculations. Correct simulation model control parameters are the basis for studying the characteristics of new energy power grids.

[0004] At present, the control parameter identification method is mainly aimed at doubly fed wind turbines. However, the grid-type doubly fed wind turbines have active and reactive power control loops, and the voltage and current inner loops are cascaded. Compared with the doubly fed wind turbines, the grid-type control parameters are more and the coupling relationship is more complex. Therefore, it is urgent to study the control parameter identification method of the grid-type doubly fed wind turbines. Summary of the invention

[0005] In order to solve the problem of how to accurately identify the control parameters of the machine-side system of a grid-type doubly-fed wind turbine generator set, the present invention provides a method for identifying partition parameters of a grid-type doubly-fed wind turbine generator set.

[0006] The grid-type doubly-fed wind turbine generator group partition parameter identification method of the present invention comprises:

[0007] S1. According to the grid-type doubly-fed wind turbine system structure, a grid-type doubly-fed wind turbine transient model is constructed, and the target error function of the grid-type doubly-fed wind turbine transient model is obtained;

[0008] S2. Partition the output data of the grid-type doubly-fed wind turbine transient model about the original machine-side control parameters to obtain multiple intervals, increase the value of the original machine-side control parameters, calculate the distance correlation coefficient of all machine-side control parameters before and after the increase about each interval, and select the highest correlation interval as the identification interval of the parameter to be identified according to the distance correlation coefficient;

[0009] S3. Taking the target error function value as the fitness, the differential evolution algorithm is used to search for the optimal solution in the identification interval. The optimization result is the machine-side control parameter in the transient model of the grid-type doubly fed wind turbine.

[0010] As a preferred embodiment, in S1, the target error function of the grid-type doubly-fed wind turbine transient model is:

[0011]

[0012] Among them, (P0, Q0) is the active and reactive power output of the grid-connected point in the transient model of the grid-type doubly fed wind turbine under the original machine-side control parameters, (P1, Q1) is the active power and reactive power output of the grid-connected point in the transient model of the grid-type doubly fed wind turbine under the identified machine-side control parameters; E represents the target error function value, and n is the total number of output data points.

[0013] Preferably, in S2, the distance correlation coefficient is:

[0014]

[0015] Among them, P2 and P3 are the active power time series of the outlet side in the interval in the transient model of the grid-type double-fed wind turbine with the original machine-side control parameters and the machine-side parameters increased, respectively. The distance covariance of the time series P2 and P3 is are the distance matrices of time series P2 and P3 after double centering, respectively. Subscripts i and j represent the positions of data points in the time series, and n is the total number of output data points;

[0016] Distance variance of time series P2

[0017] Distance variance of time series P3

[0018] As a preference,

[0019]

[0020] As a preferred embodiment, the control parameters on the generator side include the damping coefficient D and the time constant T of the active frequency controller of the power outer loop. j , the proportional coefficient k of the reactive voltage controller pv and the integration coefficient k iv , the dq axis proportional coefficient K of the voltage inner loop upd , K upq , dq axis integration coefficient K uid , K uiq , the dq axis proportional coefficient K of the current inner loop ipd , K ipq , dq axis integration coefficient K iid , K iiq .

[0021] Preferably, S3 includes:

[0022] S31. Generate initial population:

[0023]

[0024] in, are the machine-side control parameters in each iteration, the subscript i represents the population number, i=1,2,...,N, N is the population size, the superscript t represents the generation of the population, t=0 represents the initial generation, U and L are the upper and lower bounds of the machine-side control parameters, respectively; rand(0,1) is a random number between [0,1];

[0025] S32. For the individuals in the population, three different individuals are randomly selected and mutated to generate mutant individuals:

[0026]

[0027] in, represents the states of the three different individuals selected in the tth generation, is the variation term, F is the variation factor, represents the mutant individual of the i-th individual in the t-th generation;

[0028] S33. For each dimension of the individual, a crossover operation is performed according to the crossover probability CR to generate a test individual:

[0029]

[0030] in, The test subject The value of the jth dimension of A mutant individual The value of the jth dimension of Is an individual The value of the j-th dimension, CR∈[0.7,0.9];

[0031] S34. Determine the test individuals Whether to become an individual of the next generation population:

[0032] Will and The corresponding two sets of machine-side control parameters are substituted into the grid-type doubly fed wind turbine transient model, and the target error function is calculated in the identification interval. The target error function value is used as the experimental individual The fitness of the test individual The fitness is better than that of individual The fitness of replace Get the next generation population:

[0033]

[0034] Where E(·) represents the target error function;

[0035] S32-S34 are repeatedly executed until a predetermined number of iterations is completed or the iteration is completed, then the iteration is stopped, and the global optimal solution is found in the finally obtained population as the optimization result.

[0036] Preferably, the method further comprises:

[0037] S4. Compare the output data characteristics of the grid-type doubly fed wind turbine transient model under the machine-side control parameters in the optimization results and the original machine-side control parameters to evaluate the fitting degree of the optimization results:

[0038] Comprehensive determination coefficient

[0039]

[0040] (P0, Q0) is the active and reactive power output at the grid connection point in the transient model of the grid-connected double-fed wind turbine under the original machine-side control parameters, It represents the average value of active power P0 in the whole time period. It represents the average value of reactive power Q0 in the whole time period, P4 and Q4 are respectively the active power and reactive power outputted from the grid connection point in the transient model of grid-connected doubly-fed wind turbine under the machine-side control parameters in the optimization results.

[0041] The beneficial effects of the present invention are that the present invention has global search capabilities and can solve the problems of slow convergence speed, long iteration time, and complex calculation process in existing traditional optimization algorithms. The strategy of identifying control parameters based on the results of partition calculation distance correlation coefficient can significantly improve the speed and accuracy of parameter identification. The machine-side control parameters can be identified without affecting the operation process of the wind turbine set. The simulation model is repeatedly run through the m file in Matlab, and parameter identification is performed based on the output data obtained from each operation. The objective function error value of each parameter identification result is compared, and the control parameter value is continuously adjusted to significantly improve the identification accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a schematic diagram of a method for identifying control parameters of a grid-type doubly-fed wind turbine generator set on the side of the present invention;

[0043] Figure 2 It is a flowchart of the differential evolution algorithm steps of the present invention;

[0044] Figure 3 It is a schematic diagram comparing the output results of the parameter identification model of the present invention and the original model under transient conditions, including active power, reactive power, terminal voltage and terminal current at the outlet of the grid-type double-fed wind turbine generator set;

[0045] Figure 4 It is a schematic diagram of the transient condition zoning of the grid-type doubly-fed wind turbine generator system of the present invention. DETAILED DESCRIPTION

[0046] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0047] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.

[0048] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments, but they are not intended to limit the present invention.

[0049] The grid-type doubly-fed wind turbine group partition parameter identification method of this embodiment includes:

[0050] Step 1: According to the grid-type doubly-fed wind turbine system structure, a grid-type doubly-fed wind turbine transient model is built in Matlab / Simulink, the parameters of the grid-type doubly-fed wind turbine unit side controller are changed, multiple sets of output data are obtained, and the target error function of the original model and the identification model is obtained:

[0051] The parameters of the grid-connected doubly-fed wind turbine transient model are set based on the original control parameters, and the grid-connected doubly-fed wind turbine transient model is operated to obtain the active power P0 and reactive power Q0 output at its grid connection point.

[0052] By changing the machine-side parameters in the transient model of the grid-type doubly-fed wind turbine and running the transient model of the grid-type doubly-fed wind turbine, the active power P1 and reactive power Q1 of the grid-connected point of the grid-type doubly-fed wind turbine unit under the influence of the identification parameters can be obtained;

[0053] According to the target error function formula, the fitness of the selected parameter identification value in the identification process can be calculated:

[0054]

[0055] Where n is the total number of output data points.

[0056] Step 2: partition the output data of the grid-type doubly-fed wind turbine transient model about the original machine-side control parameters to obtain multiple intervals. In this step, the overall output data is divided into three intervals based on the transient curves of the active power and reactive power output at the grid-connected point of the grid-type doubly-fed wind turbine. The value of the original machine-side control parameter is increased, and the distance correlation coefficient of all machine-side control parameters before and after the increase is calculated for each interval. The highest correlation interval is selected as the identification interval of the parameter to be identified according to the distance correlation coefficient:

[0057] There are 12 machine-side control parameters for the grid-type doubly-fed wind turbine, which are the damping coefficient D and time constant T of the active frequency controller of the power outer loop. j , the proportional coefficient k of the reactive voltage controller pv and the integration coefficient k iv , the dq axis proportional coefficient K of the voltage inner loop upd K upq , dq axis integration coefficient K uid K uiq , the dq axis proportional coefficient K of the current inner loop ipd K ipq dq axis integration coefficient K iid K iiq .

[0058] The degree of correlation between each control parameter and the response curve of P and Q in the corresponding fault interval can be expressed by the distance correlation coefficient dcor. The active power time series of the outlet side in the interval in the transient model of the grid-type doubly fed wind turbine after the original machine-side control parameters and the machine-side parameters are increased are P2 and P3. The calculation formula of the distance correlation coefficient is as follows:

[0059]

[0060] Assume that all parameters increase their original values ​​by σ = 1% as the consistent evaluation standard, that is, P3 is the active output characteristic of the grid connection point after the control parameters are increased. For the above given active power output data set, the Euclidean distance between each pair of observations is first calculated to form a distance matrix. The distance matrix is ​​calculated as follows:

[0061]

[0062] For two time series P2 and P3, each time series contains n observations, and its subscript i represents the location of the data point. Based on the distance calculated above, two distance matrices A and B are formed, where A ij =d(p 2i ,p 3j ) and B ij =d(p 3i ,p 3j). The distance matrices A and B are double-centered, and the calculation formula is as follows:

[0063]

[0064] Use the doubly centered distance matrix to calculate the distance covariance and distance variance:

[0065]

[0066] Distance variance of time series P2

[0067] Distance variance of time series P3

[0068] Considering the overall transient characteristics of active power and reactive power, the distance correlation coefficients of each transient interval of active power and reactive power are added to obtain the distance correlation coefficient results of each control parameter for all transient intervals. For each control parameter, the transient interval with the maximum distance correlation coefficient is selected as the input of the optimization algorithm.

[0069] Step 3: Taking the target error function value as the fitness, the differential evolution algorithm is used to search for the optimal solution in the identification interval. The optimal solution is the machine-side control parameter in the transient model of the grid-type doubly fed wind turbine.

[0070] Step 31: Initialize the population. Randomly generate N individuals as the initial population. Each individual represents a potential solution. The initialization formula is as follows:

[0071]

[0072] in, are the machine-side control parameters in each iteration process, the subscript i represents the population number, i=1,2,...,N, N is the population size, the superscript t represents the generation of the population, t=0 represents the initial generation, U and L are the upper and lower bounds of the machine-side control parameters respectively; rand(0,1) is a random number between [0,1].

[0073] Step 32: For each individual in the population, three different individuals are randomly selected, their difference is calculated, and the difference is multiplied by a scaling factor (difference weight) F, and added to another target individual to generate a new individual (mutated individual). The basic mutation operation is as follows:

[0074]

[0075] in, represents the state of the i-th individual in the t-th generation, represents the states of three randomly selected individuals in the tth generation, which are used to generate mutant individuals. is the mutation term, and F is the mutation factor, which is used to control the scale of the mutation vector and affect the search range of the algorithm. represents the mutant individual of the i-th individual in the t-th generation.

[0076] Step 33: For each dimension of the individual, part of the genes of the mutant individual and the target individual are exchanged with a certain probability (crossover probability CR) to generate a test individual. The crossover operation formula is as follows:

[0077]

[0078] in: The test subject The value of the jth dimension of is the mutation vector The value of the jth dimension of Is an individual The value of the jth dimension, CR is the crossover probability, CR∈[0.7,0.9]. For each dimension j, if the random number is less than the crossover probability CR or equal to the randomly selected dimension j rand , the value of the j-th dimension of the test individual adopts the value of the mutation vector; otherwise, the value of the original individual is adopted. In this way, the test individual has a certain probability of inheriting the value of the mutation vector in each dimension, thereby increasing the diversity of the population.

[0079] Step 34: To determine the test individuals Becoming a member of the next generation and The corresponding two sets of machine-side control parameters are substituted into the grid-type doubly fed wind turbine transient model, and the target error function is calculated in the identification interval. The target error function value is used as the experimental individual If the fitness of the new individual is better than that of the old individual, it will be replaced:

[0080]

[0081] Steps 32 to 34 are repeatedly performed until a predetermined number of iterations is completed or the iteration is completed, then the iteration is stopped, and the global optimal solution is found in the finally obtained population as the optimization result.

[0082] The export output characteristics of the identified model obtained by the optimization algorithm and the original model are compared, and the comprehensive determination coefficient R is used to 2 To evaluate the degree of fit of the optimization result, the formula is as follows:

[0083]

[0084] Coefficient of determination R2 The value of is between 0 and 1. The closer the value is to 1, the better the fitting degree of the parameter identification result to the original model is, and the model is reliable. The closer the value is to 1, the worse the fitting effect of the model is, that is, the parameter identification result is wrong, and go to step 2 for adjustment and correction.

[0085] This implementation method collects the machine-side control parameters of the grid-type doubly-fed wind turbine under the grid voltage drop and the output active and reactive waveforms at the grid connection point, partitions the output waveform, and uses the distance correlation coefficient method to calculate the correlation coefficients between each control parameter and the interval waveform for the output waveforms in different intervals after the above partitioning. According to the correlation coefficient calculation results, the dominant area with high correlation of each control parameter is selected, and the output waveform of the high correlation area corresponding to the parameter is used for parameter identification. The Euclidean distance is selected as the objective function of the identification algorithm, and the differential evolution method is used to optimize the objective function to obtain the final parameter identification result. This method improves the overall parameter identification speed and accuracy, solves the problem of excessive error in the parallel parameter identification of the grid-type doubly-fed wind turbine, and has stronger robustness and global optimization capabilities compared with other algorithms.

[0086] Although the present invention is described herein with reference to specific embodiments, it should be understood that these embodiments are merely examples of the principles and applications of the present invention. It should therefore be understood that many modifications may be made to the exemplary embodiments and that other arrangements may be devised without departing from the spirit and scope of the present invention as defined by the appended claims. It should be understood that the various dependent claims and features described herein may be combined in a manner different from that described in the original claims. It should also be understood that features described in conjunction with individual embodiments may be used in other described embodiments.

Claims

1. A method for identifying the partition parameters of a grid-type doubly-fed wind turbine generator system, characterized in that: The method comprises: S1. According to the grid-type doubly-fed wind turbine system structure, a grid-type doubly-fed wind turbine transient model is constructed, and the target error function of the grid-type doubly-fed wind turbine transient model is obtained; S2. Partition the output data of the grid-type doubly-fed wind turbine transient model about the original machine-side control parameters to obtain multiple intervals, increase the value of the original machine-side control parameters, calculate the distance correlation coefficient of all machine-side control parameters before and after the increase about each interval, and select the highest correlation interval as the identification interval of the parameter to be identified according to the distance correlation coefficient; S3. Taking the target error function value as the fitness, the differential evolution algorithm is used to search for the optimal solution in the identification interval. The optimization result is the machine-side control parameter in the transient model of the grid-type doubly fed wind turbine.

2. The method for identifying partition parameters of a grid-type doubly-fed wind turbine generator system according to claim 1 is characterized in that: In S1, the objective error function of the grid-type doubly fed wind turbine transient model is: Among them, (P0, Q0) is the active and reactive power output of the grid-connected point in the transient model of the grid-type doubly fed wind turbine under the original machine-side control parameters, (P1, Q1) is the active power and reactive power output of the grid-connected point in the transient model of the grid-type doubly fed wind turbine under the identified machine-side control parameters; E represents the target error function value, and n is the total number of output data points.

3. The method for identifying partition parameters of a grid-type doubly-fed wind turbine generator system according to claim 1 is characterized in that: In S2, the distance correlation coefficient is: Among them, P2 and P3 are the active power time series of the outlet side in the interval in the transient model of the grid-type double-fed wind turbine with the original machine-side control parameters and the machine-side parameters increased, respectively. The distance covariance of the time series P2 and P3 is are the distance matrices of time series P2 and P3 after double centering, respectively. Subscripts i and j represent the positions of data points in the time series, and n is the total number of output data points; Distance variance of time series P2 Distance variance of time series P3 4. The method for identifying partition parameters of a grid-type doubly-fed wind turbine generator system according to claim 3 is characterized in that:

5. The method for identifying partition parameters of a grid-type doubly-fed wind turbine generator system according to claim 3 is characterized in that: A set of machine-side control parameters includes the damping coefficient D and time constant T of the active frequency controller of the power outer loop j , the proportional coefficient k of the reactive voltage controller pv and the integration coefficient k iv , the dq axis proportional coefficient K of the voltage inner loop upd , K upq , dq axis integration coefficient K uid , K uiq , the dq axis proportional coefficient K of the current inner loop ipd , K ipq , dq axis integration coefficient K iid , K iiq .

6. The method for identifying partition parameters of a grid-type doubly-fed wind turbine generator system according to claim 3 is characterized in that: S3 include: S31. Generate initial population: in, are the machine-side control parameters in each iteration, the subscript i represents the population number, i=1,2,...,N, N is the population size, the superscript t represents the generation of the population, t=0 represents the initial generation, U and L are the upper and lower bounds of the machine-side control parameters, respectively; rand(0,1) is a random number between [0,1]; S32. For the individuals in the population, three different individuals are randomly selected and mutated to generate mutant individuals: in, represents the states of the three different individuals selected in the tth generation, is the variation term, F is the variation factor, represents the mutant individual of the i-th individual in the t-th generation; S33. For each dimension of the individual, a crossover operation is performed according to the crossover probability CR to generate a test individual: in, The test subject The value of the jth dimension of A mutant individual The value of the jth dimension of Is an individual The value of the j-th dimension, CR∈[0.7,0.9]; S34. Determine the test individuals Whether to become an individual of the next generation population: Will and The corresponding two sets of machine-side control parameters are substituted into the grid-type doubly fed wind turbine transient model, and the target error function is calculated in the identification interval. The target error function value is used as the experimental individual The fitness of the test individual The fitness is better than that of individual The fitness of replace Get the next generation population: Where E(·) represents the target error function; S32-S34 are repeatedly executed until a predetermined number of iterations is completed or the iteration is completed, then the iteration is stopped, and the global optimal solution is found in the finally obtained population as the optimization result.

7. The method for identifying partition parameters of a grid-type doubly-fed wind turbine generator system according to claim 3 is characterized in that: The method further comprises: S4. Compare the output data characteristics of the grid-type doubly fed wind turbine transient model under the machine-side control parameters in the optimization results and the original machine-side control parameters to evaluate the fitting degree of the optimization results: Comprehensive determination coefficient is the active and reactive power output from the grid connection point in the transient model of the grid-connected double-fed wind turbine under the original machine-side control parameters, It represents the average value of active power P0 in the whole time period. It represents the average value of reactive power Q0 in the whole time period, P4 and Q4 are respectively the active power and reactive power outputted from the grid connection point in the transient model of grid-connected doubly-fed wind turbine under the machine-side control parameters in the optimization results.

8. A computer-readable storage device storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method for identifying partition parameters of a grid-type doubly-fed wind turbine group as claimed in any one of claims 1 to 7 are implemented.

9. A device for identifying partition parameters of a grid-type doubly-fed wind turbine group, comprising a storage device, a processor, and a computer program stored in the storage device and executable on the processor, characterized in that: The processor executes the computer program to implement the steps of the method for identifying partition parameters of a grid-type doubly-fed wind turbine group according to any one of claims 1 to 7.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method for identifying partition parameters of a grid-type doubly-fed wind turbine group as claimed in any one of claims 1 to 7 are implemented.

Images