Hydro-generator excitation system control method giving consideration to voltage response quality and low-frequency oscillation damping

Abstract

A hydro-generator excitation system control method considering voltage response quality and low-frequency oscillation damping completes excitation system PID parameter optimization and coordinated power system stabilizer parameter design in a multi-working-condition operation scene by introducing a differential evolution algorithm so as to improve the voltage regulation speed of an excitation system, reduce overshoot and improve the stability of the excitation system. The optimization objectives of shortening the adjusting time and improving the damping torque provided by the excitation system are achieved, the low-frequency-band damping of the system is improved while the voltage adjusting performance of the excitation system of the hydro-generator is improved, and low-frequency oscillation of the hydro-generator system is restrained. A power system stabilizer is introduced, the damping characteristic of the excitation system in a low-frequency band is optimized, a differential evolution algorithm is utilized, voltage regulation PID parameters of the excitation system are optimized, and PID parameter optimization of the excitation system and PSS design parameter optimization are completed. According to the method, the voltage regulation speed of the excitation system under different operation conditions can be improved, the damping of the excitation system in a low-frequency band is improved, and the response performance of the excitation system of the hydraulic generator is optimized.

Description

Technical Field

[0001] The present invention belongs to the technical field of excitation control of hydro-generators in water conservancy projects, and relates to a control method for an excitation system of a hydro-generator that takes into account both voltage response quality and low-frequency oscillation damping. Background Technique

[0002] The energy structure on the power generation side is moving forward towards the goals of "clean, low-carbon, and efficient" with high quality and high efficiency. The power system is transforming from a traditional power generation structure dominated by thermal power to a structure of multi-source collaborative power supply of thermal-hydro-wind-solar-energy storage. Among them, new energy power generation forms represented by hydropower are developing rapidly. In power grids dominated by hydropower such as the Yunnan Power Grid, hydro-generator sets have a large proportion on the power generation side and need to drive large-capacity loads through long-distance transmission lines. The power grid is prone to low-frequency oscillation phenomena, and in severe cases, it may even lead to system disconnection. The reason for the low-frequency oscillation in power grids dominated by hydropower is that the regulation performance of the excitation regulation system is different under different operating conditions, and the adaptability of the excitation system to changes in operating conditions is poor.

[0003] The regulation performance of the traditional excitation system of a hydro-generator with fixed PID parameters can meet the requirements of conventional operating conditions. However, in the face of the multi-condition operation scenario of the hydro-generator regulation system, the traditional regulation method can no longer meet the requirements.

[0004] The PID parameters of the voltage regulation of the excitation system have poor adaptability to different operating conditions. Conventional PID control has strong robustness and is easy to implement, but its parameter tuning depends on engineering experience. In the face of the current situation of multi-condition operation of the system, conventional PID control cannot ensure that the control performance always remains optimal. Therefore, it is imperative to optimize the PID parameters to improve the PID control performance. Designing PSS parameters for different operating conditions is still a difficult problem. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a control method for an excitation system of a hydro-generator that takes into account both voltage response quality and low-frequency oscillation damping. Through a differential optimization algorithm and a power system stabilizer, the adaptability of the PID parameters of the excitation regulation system under different operating conditions is improved, the voltage regulation speed is increased, the response time is shortened, the overshoot is reduced, and the damping characteristics of the excitation system in the low-frequency band are improved to suppress the system low-frequency oscillation caused by the excitation system.

[0006] To solve the above technical problem, the technical solution adopted by the present invention is: a control method for an excitation system of a hydro-generator that takes into account both voltage response quality and low-frequency oscillation damping, including the following steps: S1, for a specific power network and the corresponding excitation system of the hydro-generator it contains, select 3 to 5 representative load scenarios as the operating conditions to be optimized; S2. Calculate the specific parameter values of K2 - K6 in the excitation system according to different operating conditions of the system, and generate the excitation system parameter packages corresponding to each operating condition. S3. Based on the above excitation system and the corresponding parameter packages under different operating conditions, determine the objective function. S4. For improving the damping in the low - frequency band [0.2Hz, 2.5Hz], perform two - layer data optimization design successively. S5. Optimize the differential evolution algorithm based on minimizing the objective function. S6. Minimize the objective function when judging the optimization results.

[0007] In S4, the first - layer optimization: First, initialize the PID parameters of the voltage regulation part of the excitation system, which are given according to engineering operation experience. , The parameter tuning range; Initialize the PID parameters respectively in their respective dimensions according to the given parameter tuning range. Each individual can be represented as: ;(4) Among them, represents the number of the individual in the population, represents the generation number of evolution, represents the population size. The population size needs to be greater than or equal to 4 to facilitate subsequent mutation and crossover optimization operations of the parameters.

[0008] In S5, in the differential evolution algorithm, it is generally assumed that all randomly initialized populations conform to a uniform distribution; the parameter bounds and initial values of the PID are specified as follows: ;(5) Among them, , , represent the PID parameters for each parameter optimization; , , are the value ranges of the PID parameters respectively; represents a uniformly distributed real number randomly generated between [0, 1]. Ensure that a random initial value within the parameter bounds is generated during parameter initialization. , , are the initial values of the corresponding PID parameters after initialization respectively.

[0009] In S5, after completing the initialization of the PID parameters, perform the parameter mutation operation; randomly select three different individuals s1, s2, s3 in the decision space. For each objective , , The mutation vector of the differential evolution algorithm is as follows: ; (6) wherein, it is required that the randomly selected individual numbers s1, s2, s3 are different from each other and also different from the target number ; the mutation operator is a real constant factor used to control the scaling of the deviation variable. Therefore, during the parameter mutation process, each parameter can vary controllably within its respective bounds.

[0010] In S5, for each individual, if a solution outside the bounds appears during the mutation process, there are multiple ways to handle it. A commonly used method is boundary absorption, that is, setting the variable value exceeding the boundary to the boundary value; here, another solution is provided, taking as an example, that is or , then the following processing is carried out: ; (7) The above processing method aims to limit the mutation result within the feasible region and increase the parameter variation degree.

[0011] In S5, after parameter mutation, in order to increase the diversity of the parameter vector, a crossover operation is introduced; taking the parameter as an example, for each individual and the mutation vector , a crossover operation is carried out according to the crossover probability CR to generate a crossover vector; there are many methods for the crossover operation, and a relatively common method is binomial crossover, and its judgment formula is as follows: ; (8) wherein, represents the crossover probability, and its value range is [0, 1]; represents the jth estimated value of the random number generator generating a random number between [0, 1]; represents a randomly selected sequence, and this sequence must be between [1, D], so there must be a ; in binomial crossover, for each dimension j, if the condition one is satisfied, the crossover vector is the mutation vector, and if the condition two is satisfied, the crossover vector is the individual value.

[0012] In S5, after the above mutation and crossover operations, the individual PID parameters and the crossover vector PID parameters are respectively set in the excitation system. According to the greedy criterion, the effects of the two groups of parameters are compared. If the effect of the crossover vector is better than that of the individual parameters, the crossover vector is retained as the individual of the next generation, otherwise the individual parameters are retained; After each mutation and crossover, it is judged whether the objective function meets the conditions; for the objective function, reducing the adjustment time will inevitably sacrifice the damping characteristics, while reducing the overshoot requires improving the damping performance, and the two objectives are contradictory; therefore, a weight coefficient is introduced to meet the optimization requirements. and , which are used to form the objective function: ; (9) where os represents the overshoot and st represents the adjustment time, and ; In the optimization process, the weighted objective function should be made smaller. If the optimization objective requirements are met, the optimal PID parameters are output, or if the maximum number of iterations is reached, the optimal PID parameter values corresponding to the minimum of the objective function G are output.

[0013] In S4, the second-layer optimization: The second part of the parameter optimization of the power system stabilizer is based on the first-layer PID parameter optimization; in the case of not adding the power system stabilizer, the excitation system transfer function is established for the excitation system parameters and the corresponding PID optimal parameters under different operating conditions, and the torque method decomposition is carried out on the excitation system transfer function: ; (10) where represents the original transfer function of the excitation system corresponding to the th operating condition, and respectively represent the initial values of the damping coefficient and synchronous coefficient provided by the excitation system under the th operating condition; After introducing the optimized control of the power system stabilizer, the transfer function of the excitation system is: ; (11) where and respectively represent the damping coefficient and synchronous coefficient provided by the excitation system after introducing the optimized control.

[0014] In S5, the objective function of the differential evolution algorithm in the second-layer optimization is that the damping coefficient of the excitation system after parameter optimization is greater than the original damping coefficient in the low-frequency band, and the damping is maximized as much as possible; The optimization objects are the gain coefficient of the power system stabilizer and the time constant of the lead-lag link; first, the number of lead-lag link stages is determined, and then the basic structure of the power system stabilizer is determined, and then the parameter optimization is carried out; First, initialize the data, limit the value ranges of the gain coefficient and the time constant of the lead-lag link, and then generate the initial parameters of the PSS through a random mode; perform mutation and crossover based on the original parameters to generate a crossover vector; Establish the objective function: ; (12) Among them, represents the damping ratio of the characteristic roots corresponding to the system in the low-frequency band; That is, the goal of the second-layer parameter optimization is that among all the characteristic roots corresponding to the optimized system in the low-frequency band, the damping ratio of the characteristic root with the smallest damping ratio is the largest, and after taking the negative value, the value of the objective function is the smallest.

[0015] In S6, after each data optimization, substitute the PSS parameters into the transfer function to determine whether the parameters make the damping coefficient greater than the original damping coefficient and make the objective function the smallest; if the objective function condition is met or the maximum number of iterations is reached, output the optimal PSS parameters.

[0016] The main beneficial effects of the present invention are as follows: By introducing a power system stabilizer, the damping characteristics of the system in the low-frequency band are improved, and a differential evolution algorithm is introduced to optimize the PID parameters of the voltage regulation part of the excitation system and the PSS parameters. By optimizing the PID parameters, the response speed of the voltage regulation process of the hydrogenerator excitation system can be improved, the voltage response regulation time can be reduced, and the overshoot can be reduced.

[0017] By optimizing the PSS parameters, the damping performance of the excitation system in the low-frequency band can be further improved, thereby improving the damping characteristics of the excitation system in the low-frequency band, increasing the system damping ratio, and suppressing the low-frequency oscillation phenomenon caused by the poor performance of the excitation system in multiple operating conditions in a high-ratio hydrogenerator system. Description of the Drawings

[0018] The present invention will be further described below with reference to the drawings and embodiments.

[0019] Figure 1 It is a Philips-Heffron model diagram of a single-machine infinite system.

[0020] Figure 2 It is a flowchart of the differential evolution algorithm for optimizing the parameters of the excitation system of the present invention.

[0021] Figure 3 It is a damping torque vector diagram of the present invention.

[0022] Figure 4 It is a control block diagram of the damping controller of the present invention. Specific Embodiments

[0023] As Figures 1 to 4In a hydrogenerator excitation system control method that takes into account both voltage response quality and low-frequency oscillation damping, Example 1. Analysis of the existing excitation system: Changes in the system operating conditions result in the excitation system providing negative damping. For example, Figure 1 The Philips-Heffron model of a single-machine infinite-bus system is shown as follows. Loop 1 represents the rotor motion equation of the generator. Loop 1 mainly controls the stability of the generator power angle to achieve power balance. Loop 2 represents the voltage regulator and the excitation system equation. Loop 2 mainly controls the stability of the generator terminal voltage and assists in regulating the system reactive power balance. The magnitudes of the parameters K2 to K6 in Loop 2 are related to the system structure and operating conditions, and the corresponding formulas for each parameter are as follows ;(1) In the formula, and are the q-axis and d-axis reactances of the generator respectively; is the transient reactance of the generator; and are the initial values of the q-axis component and d-axis component of the stator voltage respectively; is the power angle; is the voltage at the connection point; is the line impedance; is the imaginary potential of the generator.

[0024] Generally, the values of K2 to K4 and K6 are positive. During heavy load operation, the system power angle difference is large, and the excitation system parameter K5 may be negative. The change in the sign of this parameter has an important impact on the system low-frequency oscillation mode. During the system dynamic response process, the excitation system may exhibit negative damping characteristics in the low-frequency band, resulting in a reduction in the system low-frequency band damping and obvious low-frequency oscillations.

[0025] To address the phenomenon of low damping or negative damping in the power system, a power system stabilizer (PSS) is usually introduced for optimal control. The PSS consists of an amplification link, a signal filtering link, a lead-lag link, and a limiting link. Its input signals are mainly the speed deviation, frequency deviation, electromagnetic power deviation, and mechanical power deviation, and the output signal is superimposed at the location to be compensated. The PSS formula is as follows: (2) In the formula,; is the PSS gain coefficient; and are the time constants of the lead-lag link, is the DC blocking constant.

[0026] According to the extreme value theorem (3) Take Substitute , and we get

[0027] It can be seen that the power system has no effect in the stable operation state and only participates in the system dynamic response process. Through parameter design, PSS can provide positive damping for the system, thereby improving the system damping characteristics and optimizing the response performance. However, the PSS parameter design for different operating conditions remains a difficult problem.

[0028] For the PID parameter optimization of the voltage regulation module and the PSS parameter design optimization, the differential evolution algorithm can basically meet the requirements. The differential evolution algorithm is a global optimization algorithm suitable for continuous optimization problems. By simulating the mechanisms of natural selection and genetic variation, it can effectively find the global optimal solution in a complex search space. Through data mutation and crossover operations, it can effectively search the entire space and avoid falling into local optimal solutions. Moreover, it has less dependence on the initial population. Even if the quality of the initial population is not high, it can still find the global optimal solution. For the problem that the regulation performance of the excitation system deteriorates due to the change of the operating conditions of the hydro-generator, the differential evolution algorithm can parallelly process the PID parameter optimization and the PSS parameter optimization according to the basic operating parameters corresponding to different operating conditions and multiple optimization constraints, and is suitable for multi-constraint optimization problems of optimizing the response speed, overshoot, and damping characteristics.

[0029] Based on this, the present invention proposes a control method for the excitation system of a hydro-generator that takes into account both the voltage response quality and the low-frequency oscillation damping. Specifically, a power system stabilizer is introduced to optimize the damping characteristics of the excitation system in the low-frequency band, and the differential evolution algorithm is used to optimize the voltage regulation PID parameters of the excitation system, completing the PID parameter optimization of the excitation system and the PSS design parameter optimization. This method can improve the voltage regulation speed of the excitation system under different operating conditions, increase the damping of the excitation system in the low-frequency band, and optimize the response performance of the excitation system of the hydro-generator.

[0030] Embodiment 2 By introducing a power system stabilizer and the differential evolution algorithm, it is possible to specifically improve the damping characteristics of the system in the low-frequency band and complete the PID parameter optimization of the voltage regulation module and the PSS parameter design optimization under different system operating conditions. Facing different operating conditions, it can improve the voltage regulation response speed of the excitation system, reduce the overshoot in the voltage response process, and specifically complete the low-frequency band damping compensation through the parameter design optimization of the power system stabilizer, improve the damping performance of the excitation system in the low-frequency band, and further suppress the system low-frequency oscillation.

[0031] Specifically: For a specific power grid and the excitation system corresponding to the hydro-generators it contains, select 3 to 5 representative load scenarios as the operating conditions to be optimized. Calculate the specific K2 - K6 parameter values in the excitation system according to different operating conditions of the system, and generate the parameter packages of the excitation system corresponding to each operating condition. Based on the above excitation system and the corresponding parameter packages under different operating conditions, determine the objective function: compared with the original parameter settings of the excitation system, the response time is reduced by 20%, the overshoot is reduced by 15%, and the damping in the low-frequency band [0.2 Hz, 2.5 Hz] is increased. Conduct two-layer data optimization design in sequence. It should be noted that the differential evolution algorithm is optimized based on minimizing the objective function, so the objective function needs to be minimized when judging the optimization results.

[0032] The first layer of optimization: First, initialize the PID parameters of the voltage regulation part of the excitation system, and give , , the parameter setting range according to engineering operation experience. Initialize the PID parameters respectively in their respective dimensions according to the given parameter setting range. Each individual can be expressed as: ; (4) Among them, represents the number of the individual in the population, represents the generation number of evolution, represents the population size, and the population size needs to be greater than or equal to 4 to facilitate subsequent mutation and crossover optimization operations of the parameters.

[0033] In the differential evolution algorithm, it is generally assumed that all randomly initialized populations conform to a uniform distribution. The parameter bounds and initial values of the PID are specified as follows: ; (5) Among them, , , represent the PID parameters for each parameter optimization. , , are the value ranges of the PID parameters respectively. represents a uniformly distributed real number randomly generated between [0, 1]. Ensure that a random initial value within the parameter bounds is generated during parameter initialization, , , are the initial values of the corresponding PID parameters after initialization respectively.

[0034] After the PID parameter initialization is completed, the parameter mutation operation is carried out. Randomly select three different individuals s1, s2, s3 within the decision space. For each objective , , The mutation vector of the differential evolution algorithm is as follows: ;(6) Among them, it is required that the randomly selected individual numbers s1, s2, s3 are different from each other and also different from the objective number . The mutation operator is a real constant factor used to control the scaling of the deviation variable. Therefore, during the parameter mutation process, each parameter can vary controllably within its respective bounds.

[0035] For each individual, if a solution outside the bounds appears during the mutation process, there are multiple ways to handle it. A commonly used method is boundary absorption, that is, setting the variable value exceeding the boundary to the boundary value. Here, another solution is provided. Taking as an example, that is or , then the following processing is carried out: ;(7) The above processing method aims to limit the mutation result within the feasible region and increase the degree of parameter variation.

[0036] After the parameter mutation, in order to increase the diversity of the parameter vector, a crossover operation is introduced. Taking the parameter as an example, for each individual and the mutation vector , a crossover operation is carried out according to the crossover probability CR to generate a crossover vector. There are many methods for the crossover operation. A relatively common method is binomial crossover, and its judgment formula is as follows: ;(8) Among them, represents the crossover probability, and its value range is [0,1]. represents the j-th estimated value of the random number generator that generates a random number between [0,1]. represents a randomly selected sequence, and this sequence must be between [1,D], so there must be a . In binomial crossover, for each dimension j, if condition one is satisfied, the crossover vector is the mutation vector, and if condition two is satisfied, the crossover vector is the individual value.

[0037] After the above mutation and crossover operations, individual PID parameters and cross-vector PID parameters are respectively set in the excitation system. According to the greedy criterion, the effects of the two sets of parameters are compared. If the effect of the cross-vector is better than that of the individual parameters, the cross-vector is retained as the individual of the next generation; otherwise, the individual parameters are retained.

[0038] After each mutation and crossover, it is judged whether the objective function meets the conditions. For the objective function, reducing the adjustment time will inevitably sacrifice the damping characteristics, while reducing the overshoot requires improving the damping performance, and the two objectives are contradictory. Therefore, a weight coefficient and are introduced to form the objective function: ; (9) where os represents the overshoot, st represents the adjustment time, and .

[0039] In the optimization process, the weighted objective function should be made smaller. If the optimization objective requirements are met, the optimal PID parameters are output; if the maximum number of iterations is reached, the optimal PID parameter values corresponding to the minimum of the objective function G are output.

[0040] Second-layer optimization: The parameter optimization of the power system stabilizer in the second part is based on the first-layer PID parameter optimization. In the case where the power system stabilizer is not added, the excitation system transfer function is established for the excitation system parameters and the corresponding optimal PID parameters under different operating conditions, and the torque method decomposition is performed on the excitation system transfer function: ; (10) where represents the original transfer function of the excitation system corresponding to the th operating condition, and and respectively represent the initial values of the damping coefficient and synchronous coefficient provided by the excitation system under the th operating condition.

[0041] After introducing the optimized control of the power system stabilizer, the transfer function of the excitation system is: ; (11) where and respectively represent the damping coefficient and synchronous coefficient provided by the excitation system after introducing the optimized control.

[0042] The objective function of the differential evolution algorithm in the second-layer optimization is that the damping coefficient of the excitation system after parameter optimization in the low-frequency band is greater than the original damping coefficient, and the damping is maximized as much as possible.

[0043] The objects to be optimized are the gain coefficient of the power system stabilizer and the time constants of the lead-lag links. First, determine the number of stages of the lead-lag links, and then determine the basic structure of the power system stabilizer. Secondly, carry out parameter optimization.

[0044] First, perform data initialization, limit the value ranges of the gain coefficient and the time constants of the lead-lag links, and then generate the initial parameters of the PSS through a random mode. Based on the original parameters, perform mutation and crossover to generate crossover vectors.

[0045] Establish the objective function: ; (12) Among them, represents the damping ratio of the characteristic roots corresponding to the low-frequency band of the system.

[0046] That is, the objective of the second-layer parameter optimization is that among all the characteristic roots corresponding to the low-frequency band of the optimized system, the damping ratio of the characteristic root with the smallest damping ratio is the largest. After taking the negative value, the value of the objective function is the smallest.

[0047] After each data optimization, substitute the PSS parameters into the transfer function to determine whether the parameters make the damping coefficient greater than the original damping coefficient and make the objective function the smallest. If the objective function conditions are met or the maximum number of iterations is reached, output the optimal PSS parameters.

[0048] In the above method, by introducing a power system stabilizer, the damping characteristics of the system in the low-frequency band are improved, and a differential evolution algorithm is introduced to optimize the PID parameters of the voltage regulation part of the excitation system and the PSS parameters. By optimizing the PID parameters, the response speed of the voltage regulation process of the hydrogenerator excitation system can be improved, the voltage response regulation time can be reduced, the overshoot can be reduced, and by optimizing the PSS parameters, the damping performance of the excitation system in the low-frequency band can be further improved, thereby improving the damping characteristics of the excitation system in the low-frequency band, increasing the system damping ratio, and suppressing the low-frequency oscillation phenomenon caused by the poor performance of the excitation system in multiple operating conditions in a high-ratio hydrogenerator system.

[0049] The above embodiments are only the preferred technical solutions of the present invention and should not be regarded as limitations on the present invention. The embodiments and the features in the embodiments in this application can be arbitrarily combined with each other without conflict. The protection scope of the present invention should be the technical solutions recorded in the claims, including the equivalent replacement solutions of the technical features in the technical solutions recorded in the claims. That is, the equivalent replacement improvements within this scope are also within the protection scope of the present invention.

Claims

1. A control method for the excitation system of a hydro-generator that takes into account both voltage response quality and low-frequency oscillation damping, characterized in that It includes the following steps: S1. For a specific power grid and the excitation system corresponding to the hydro-generator contained therein, select 3 to 5 representative load scenarios as the operating conditions to be optimized; S2. Calculate the specific parameter values of K2 to K6 in the excitation system according to different operating conditions of the system, and generate the parameter packages of the excitation system corresponding to each operating condition; S3. Based on the above excitation system and the parameter packages corresponding to different operating conditions, determine the objective function; S4. For the improvement of damping in the low-frequency band [0.2Hz, 2.5Hz], perform two-layer data optimization design in sequence; S5. Optimize using the differential evolution algorithm based on minimizing the objective function; S6. Minimize the objective function when judging the optimization result.

2. The control method of the excitation system of a hydro-generator that takes into account both voltage response quality and low-frequency oscillation damping according to claim 1, characterized in that: In S4, the first layer of optimization: First, initialize the PID parameters of the voltage regulation part of the excitation system and give them according to engineering operation experience. , , The parameter setting range; Initialize the PID parameters respectively in their respective dimensions according to the given parameter setting range. Each individual can be represented as: ；(4) Among them, represents the individual's number in the population, represents the generation number of evolution, represents the population size, and the population size needs to be greater than or equal to 4 to facilitate subsequent mutation and crossover optimization operations on the parameters.

3. The control method for the excitation system of a hydro-generator that takes into account both voltage response quality and low-frequency oscillation damping according to claim 1, characterized in that: In S5, in the differential evolution algorithm, it is generally assumed that all randomly initialized populations conform to a uniform distribution; the parameter bounds and initial parameter values of the PID are specified as follows: ；(5) Among them, , , represent the PID parameters for each parameter optimization; , , are the value ranges of the PID parameters respectively; represents a uniform real number randomly generated between [0, 1], and ensures that a random initial value within the parameter bounds is generated during parameter initialization. , , are the initial values of the corresponding PID parameters after initialization respectively.

4. The control method for the excitation system of a hydro-generator that takes into account both voltage response quality and low-frequency oscillation damping according to claim 3, characterized in that: In S5, after completing the PID parameter initialization, a parameter mutation operation is performed; three different individuals s1, s2, and s3 are randomly selected within the decision space, and for each objective , , The mutation vectors of the differential evolution algorithm are as follows: ；(6) Among them, it is required that the randomly selected individual serial numbers s1, s2, and s3 are different from each other and also different from the target serial number ; the mutation operator is a real constant factor used to control the scaling of the deviation variable. Therefore, during the parameter mutation process, each parameter can vary controllably within its respective bounds.

5. The control method of the excitation system of a hydro-generator that takes into account both voltage response quality and low-frequency oscillation damping according to claim 4, characterized in that: In S5, for each individual, if a solution outside the boundary appears during the mutation process, there are multiple ways to handle it. A commonly used method is boundary absorption, that is, setting the variable value exceeding the boundary to the boundary value. Here, another solution is provided. Taking as an example, that is or , the following processing is carried out: ；(7) The above processing method aims to limit the mutation result within the feasible domain and increase the parameter variation degree.

6. The control method for the excitation system of a hydro-generator that takes into account both voltage response quality and low-frequency oscillation damping according to claim 5, characterized in that: at In S5, after the parameters mutate, in order to increase the diversity of the parameter vectors, a crossover operation is introduced; taking the parameter as an example, for each individual and the mutation vector , a crossover vector is generated by performing a crossover operation according to the crossover probability CR; there are many methods for the crossover operation, and the relatively common method is binomial crossover, and its judgment formula is shown as follows: ； (8) Among them, represents the crossover probability, and its value range is [0, 1]; represents the j-th estimated value of the random number generator that generates a random number between [0, 1]; represents a randomly selected sequence, and this sequence must be between [1, D], so there must be a ; in binomial crossover, for each dimension j, if condition one is satisfied, the crossover vector is the mutant vector, and if condition two is satisfied, the crossover vector is the individual value.

7. The control method of the excitation system of a hydro-generator that takes into account both voltage response quality and low-frequency oscillation damping according to claim 6, characterized in that: In S5, after the above mutation and crossover operations, set the individual PID parameters and the crossover vector PID parameters in the excitation system respectively, compare the effects of the two groups of parameters according to the greedy criterion, if the effect of the crossover vector is better than that of the individual parameters, then retain the crossover vector as the individual of the next generation, otherwise retain the individual parameters; After each mutation and crossover, it is judged whether the objective function meets the conditions; for the objective function, reducing the adjustment time will inevitably sacrifice the damping characteristics, while reducing the overshoot requires improving the damping performance, and the two objectives are contradictory; therefore, a weight coefficient is introduced to meet the optimization requirements and , which are used to form the objective function: ； (9) where os represents the overshoot and st represents the settling time, and ; The optimization process should make the weighted objective function smaller. If the optimization objective requirements are met, output the optimal PID parameters, or if the maximum number of iterations is reached, output the optimal PID parameter value corresponding to the minimum of the objective function G.

8. The control method for the excitation system of a hydro-generator that takes into account both voltage response quality and low-frequency oscillation damping according to claim 1, characterized in that: In S4, the second layer of optimization: The parameter optimization of the power system stabilizer in the second part is based on the first layer of PID parameter optimization; in the case where the power system stabilizer is not added, establish the excitation system transfer function for the excitation system parameters and the corresponding optimal PID parameters under different operating conditions, and decompose the excitation system transfer function by the torque method: ； (10) Among them, represents the original transfer function of the excitation system corresponding to the th operating condition, and respectively represent the initial values of the damping coefficient and the synchronizing coefficient provided by the excitation system under the th operating condition; After introducing the power system stabilizer for optimal control, the transfer function of the excitation system is: ；(11) Among them, and respectively represent the damping coefficient and synchronous coefficient provided by the excitation system after introducing the optimized control.

9. The control method for the excitation system of a hydro-generator that takes into account both voltage response quality and low-frequency oscillation damping according to claim 1, characterized in that: In S5, the objective function of the differential evolution algorithm in the second layer of optimization is that the damping coefficient of the excitation system after parameter optimization in the low-frequency band is greater than the original damping coefficient, and the damping is maximized as much as possible; The optimization objects are the gain coefficient of the power system stabilizer and the time constants of the lead-lag links; first determine the number of lead-lag links, and then determine the basic structure of the power system stabilizer, and then perform parameter optimization; First, perform data initialization, limit the value ranges of the gain coefficient and the time constants of the lead-lag links, and then generate the initial parameters of the PSS through a random mode; perform mutation and crossover based on the original parameters to generate a crossover vector; Establish the objective function: ； (12) Among them, represents the damping ratio of the characteristic roots corresponding to the system in the low-frequency band; That is, the objective of the second layer of parameter optimization is that the damping ratio of the eigenvalue with the smallest damping ratio among all the eigenvalues corresponding to the system in the low-frequency band after optimization is the largest, and the objective function value is the smallest after taking the negative value.

10. The excitation system control method for a hydro-generator that takes into account both voltage response quality and low-frequency oscillation damping according to claim 1, characterized in that: at In S6, after each data optimization, substitute the PSS parameters into the transfer function to judge whether the parameters make the damping coefficient greater than the original damping coefficient and make the objective function the smallest; Output the optimal PSS parameters when the objective function condition is satisfied or the maximum number of iterations is reached.

Images