TWI628901B - Method of configuring self-excited capacitors for single-phase induction generator - Google Patents

Abstract

A self-excited capacitor configuration method for a single-phase induction generator, the self-excited capacitor configuration method comprising: (a) setting a plurality of individuals; (b) determining parameters of the individuals in a random manner; (c) by using the The individual's parameters calculate the overall voltage change rate of each individual; (d) select a plurality of selected individuals and a plurality of unsuccessful individuals by the overall voltage change rate of each individual; (e) randomly exchange each of the selected individuals. One of the parameters is a plurality of selected sub-individuals, and one of the parameters randomly changing at least one of the unsuccessful individuals is at least one mutant sub-population; (f) from the selected individuals, the selected sub-individuals, and the mutations The sub-individual selects an individual to be selected; and (g) determines whether the rate of change of the voltage of the selected individual is within a range of error convergence.

Description

Self-excited capacitor configuration method for single-phase induction generator

This creation is about a self-excited capacitor configuration method for a single-phase induction generator, especially a self-excited capacitor configuration method for a single-phase induction generator using a genetic algorithm.

The conventional induction motor is divided into a single-phase induction motor and a three-phase induction motor. When the induction motor is connected to an alternating current operation of frequency f, there is no constant proportional relationship between the rotational speed n and the alternating current frequency f. There is a slip between the synchronous speed and the rotor speed, and the slip is usually between 3% and 10%. When the induction motor is added with a power such that the speed of the induction motor exceeds the synchronous speed, it becomes a generator. Induction generators have advantages that are not found in synchronous generators, such as simple construction, low cost, low maintenance costs, and no DC excitation devices. Therefore, induction generators are well suited for some smaller power generation equipment, such as wind turbines, small hydroelectric generators, and engine-driven generators. For the use of electricity in remote areas and special places, it is a good choice.

Although the self-excitation of the capacitance of the induction machine was discovered as early as 50 to 60 years ago, it is not worthy of attention because it cannot propose an effective method to control the voltage fluctuation of the load terminal. Until recently, due to the rise of Green Energy Technology, research and application of self-excited induction generators (SEIG) have continued. First, because induction generators have some advantages that are not available in synchronous generators, such as units. Low cost, rugged construction, no DC excitation, no brushes (squirrel cage rotor) and low maintenance costs. Therefore, the induction generator is used in remote rural areas as an independent power generation device with good power generation. Therefore, the current induction generator is considered to be a fairly good small power generation equipment.

Among the induction generators, the application field of single-phase induction generators is the widest. However, since the load terminal voltage generated by the single-phase induction generator is easily affected by the load current, the power generation performance of the single-phase induction generator is low. In order to solve the above situation, the conventional single-phase induction generators are supplemented with self-excited capacitors to improve the power generation performance of the single-phase induction generator. However, in the existing dual-winding self-excitation structure of the single-phase induction generator, the current method for determining the candidate capacitance of the self-excited capacitor mostly uses a trial method to determine a suitable set of capacitance candidate values (Appropriate Value). Therefore, the result of optimizing the candidate capacitance of the self-excited capacitor cannot be achieved, and the fluctuation of the voltage at the load terminal cannot be effectively reduced.

Therefore, for the single-phase induction generator, the double-winding self-excitation structure optimizes the capacitance candidate capacitance of the parallel capacitor and the series capacitor. Based on the enhanced gene algorithm, a strategy and method for improving the efficiency of induction generators is designed, which is a major problem that the creators of this case want to overcome and solve.

In order to solve the above problems, the present invention provides a self-excited capacitance configuration method of a single-phase induction generator to overcome the problems of the prior art. Therefore, in the first embodiment of the present invention, the single-phase induction generator includes a main winding and an auxiliary winding, the main winding is connected in series with a series capacitor and then connected in parallel with a load, and the auxiliary winding is connected in parallel with a parallel capacitor. The method for configuring the stimuli includes: (a) setting a plurality of individuals, each of which includes a candidate capacitance of the series capacitor and a candidate capacitance of the parallel capacitor. (b) determining the candidate capacitance of the series capacitance of the individual and the candidate capacitance of the parallel capacitance in a random manner. (c) calculating a first terminal voltage of the load of each individual by the candidate capacitance of the series capacitance of each individual and the candidate capacitance of the parallel capacitance, and calculating each individual by the first terminal voltage The overall rate of voltage change. (d) A plurality of selected individuals are selected by the overall voltage change rate of each individual, and the unselected individuals are a plurality of unsuccessful individuals. (e1) randomly selecting the candidate capacitance of the series capacitance of each selected individual or the candidate capacitance of the parallel capacitance, and generating a plurality of selected sub-individuals. (e2) randomly selecting at least one of the selected individuals to be at least one mutant individual, and randomly changing the candidate capacitance of the series capacitance or the candidate capacitance of the parallel capacitance within the individual of the mutants. (f) calculating, by step (c), a second terminal voltage of the load of each of the selected sub-individuals and each of the mutant sub-individuals, and calculating each of the selected sub-individuals and each by the second terminal voltage The overall voltage change rate of the mutant individual. (g) sorting the selected individuals, the selected sub-individuals, and the magnitude of the overall voltage change rate of the individual mutants, and selecting a selected individual having the smallest overall rate of change of the voltage, and determining the individual of the selected individual Whether the rate of voltage change is within a range of error convergence. Wherein, if the voltage change rate of the selected individual is within the error convergence range, the candidate capacitance of the series capacitor representing the selected individual and the candidate capacitance of the parallel capacitor are selected as the configuration of the self-excited capacitor.

In the first embodiment, the step (g) further includes: (g1) if the voltage change rate of the selected individual is not within the error convergence range, the selected individuals, the selected sub-individuals, and the Some of the mutant individuals are set to the individuals and jump back to step (c).

In the first embodiment, after the step (g1), the method further includes: (g2), if the step (c) to the step (g1) are repeatedly performed over a maximum number of iterations, the selected individuals are selected, and the selected ones are selected. Among the sub-individuals and the individual mutants, the closest individual is closest to the error convergence range, and the candidate capacitance of the series capacitance representing the proximity individual and the candidate capacitance of the parallel capacitance are selected.

In the first embodiment, the step (d) further includes: (d1) ordering the overall voltage change rate of each individual from large to small, and the smaller the overall voltage change rate of the individuals, the greater the gain. Selection rate.

In order to solve the above problems, the present invention provides a self-excited capacitance configuration method of a single-phase induction generator to overcome the problems of the prior art. Therefore, in the second embodiment of the present invention, the single-phase induction generator includes a main winding and an auxiliary winding, the main winding is connected in parallel with a main winding short parallel capacitor and then connected in series with a series capacitor and a load. Parallel-parallel capacitors, the self-excited capacitor configuration method includes: (a) setting a plurality of individuals, each individual including a candidate capacitance of the short-short capacitor of the main winding, a candidate capacitance of the series capacitor, and a candidate for the parallel capacitor Capacitance. (b) determining, in a random manner, the candidate capacitance of the short winding capacitance of the main winding of the individual, the candidate capacitance of the series capacitance, and the candidate capacitance of the parallel capacitance. (c) calculating, by each individual, a candidate capacitance of the short-short capacitor of the main winding, a candidate capacitance of the series capacitor, and a candidate capacitance of the parallel capacitor, calculating a first terminal voltage of the load of each individual, and The overall voltage change rate of each individual is calculated by the first terminal voltage. (d) A plurality of selected individuals are selected by the overall voltage change rate of each individual, and the unselected individuals are a plurality of unsuccessful individuals. (e1) randomly exchanging the candidate capacitance of the short winding capacitance of the main winding of each selected individual, the candidate capacitance of the series capacitance or the candidate capacitance of the parallel capacitance, and generating a plurality of selected sub-individuals. (e2) randomly selecting at least one of the selected individuals to be at least one mutant individual, and randomly changing the candidate capacitance of the short winding capacitance of the main winding in the individual of the mutant, the candidate capacitance of the series capacitance or the The candidate capacitance of the shunt capacitor. (f) calculating, by step (c), a second terminal voltage of the load of each of the selected sub-individuals and each of the mutant sub-individuals, and calculating each of the selected sub-individuals and each by the second terminal voltage The overall voltage change rate of the mutant individual. (g) sorting the selected individuals, the selected sub-individuals, and the magnitude of the overall voltage change rate of the individual mutants, and selecting a selected individual having the smallest overall rate of change of the voltage, and determining the individual of the selected individual Whether the rate of voltage change is within a range of error convergence. Wherein, if the voltage change rate of the selected individual is within the error convergence range, the candidate capacitance value of the short winding capacitor of the main winding representing the selected individual, the candidate capacitance of the series capacitor and the candidate of the parallel capacitor are selected. The capacitance is used as a configuration of the self-excited capacitor.

In the second embodiment, the step (g) further includes: (g1) if the voltage change rate of the selected individual is not within the error convergence range, the selected individuals, the selected sub-individuals, and the Some of the mutant individuals are set to the individuals and jump back to step (c).

In the second embodiment, after the step (g1), the method further includes: (g2), if the step (c) to the step (g1) are repeatedly performed over a maximum number of iterations, the selected individuals are selected, and the selected ones are selected. a sub-individual and one of the mutant sub- individuals, closest to the error convergence range, and selecting a candidate capacitance value of the short-parallel capacitance of the main winding representing the proximity individual, a candidate capacitance of the series capacitance, and the parallel capacitance Candidate value.

In the second embodiment, wherein the step (d) further comprises: (d1) ordering the overall voltage change rate of each individual from large to small, and the smaller the overall voltage change rate of the individuals, the greater the gain. Selection rate.

In order to solve the above problems, the present invention provides a self-excited capacitance configuration method of a single-phase induction generator to overcome the problems of the prior art. Therefore, in the third embodiment of the present invention, the single-phase induction generator includes a main winding and an auxiliary winding, the main winding is connected in series with a series capacitor and then a main winding long parallel capacitor and a load, the auxiliary winding system Parallel-parallel capacitors, the self-excited capacitor configuration method includes: (a) setting a plurality of individuals, each individual including a candidate capacitance of the series capacitor, a candidate capacitance of the main winding long parallel capacitor, and a candidate for the parallel capacitor Capacitance. (b) determining, in a random manner, a candidate capacitance of the series capacitance of the individual, a candidate capacitance of the main winding long parallel capacitance, and a candidate capacitance of the parallel capacitance. (c) calculating, by each individual, the candidate capacitance of the series capacitor, the candidate capacitance of the main winding long parallel capacitor, and the candidate capacitance of the parallel capacitor, calculating a first terminal voltage of the load of each individual, and The overall voltage change rate of each individual is calculated by the first terminal voltage. (d) A plurality of selected individuals are selected by the overall voltage change rate of each individual, and the unselected individuals are a plurality of unsuccessful individuals. (e1) randomly exchanging the candidate capacitance of the series capacitance of each selected individual, the candidate capacitance of the main winding long parallel capacitance or the candidate capacitance of the parallel capacitance, and generating a plurality of selected sub-individuals. (e2) randomly select the selected ones At least one of the bodies is at least one mutant individual, and randomly changes the candidate capacitance of the series capacitance within the individual of the mutant, the candidate capacitance of the main winding long parallel capacitance or the candidate capacitance of the parallel capacitance. (f) calculating, by step (c), a second terminal voltage of the load of each of the selected sub-individuals and each of the mutant sub-individuals, and calculating each of the selected sub-individuals and each by the second terminal voltage The overall voltage change rate of the mutant individual. (g) sorting the selected individuals, the selected sub-individuals, and the magnitude of the overall voltage change rate of the individual mutants, and selecting a selected individual having the smallest overall rate of change of the voltage, and determining the individual of the selected individual Whether the rate of voltage change is within a range of error convergence. Wherein, if the voltage change rate of the selected individual is within the error convergence range, the candidate capacitance of the series capacitor representing the selected individual, the candidate capacitance of the long winding capacitance of the main winding, and the candidate of the parallel capacitance are selected. The capacitance is used as a configuration of the self-excited capacitor.

In the third embodiment, the step (g) further includes: (g1) if the voltage change rate of the selected individual is not within the error convergence range, the selected individuals, the selected sub-individuals, and the Some of the mutant individuals are set to the individuals and jump back to step (c).

In the third embodiment, after the step (g1), the method further includes: (g2) if the step (c) to the step (g1) are repeatedly performed over a maximum number of iterations, the selected individuals are selected, and the selected ones are selected. a sub-individual and one of the mutant sub- individuals, closest to the approximate convergence range of the error, and selecting a candidate capacitance of the series capacitance representing the proximity individual, a candidate capacitance of the main winding long parallel capacitance, and the parallel capacitance Candidate value.

In the third embodiment, wherein the step (d) further comprises: (d1) ordering the overall voltage change rate of each individual from large to small, and the smaller the overall voltage change rate of the individuals, the greater the gain. Selection rate.

In order to further understand the technology, the means and the effect of the present invention in order to achieve the intended purpose, refer to the following detailed description of the invention and the accompanying drawings. Specific understanding, however, the drawings are provided for reference and explanation only, It is not intended to limit the invention.

100, 100A, 100B‧‧‧ induction generator

10, 10A, 10B‧‧‧ rotor

20, 20A, 20B‧‧‧ Stator

30, 30A, 30B‧‧‧ prime mover

21, 21A, 21B‧‧‧ main winding

22, 22A, 22B‧‧‧Auxiliary winding

23, 23A, 23B‧‧‧ load

Vo‧‧‧ terminal voltage

Vo1‧‧‧ first terminal voltage

Vo2‧‧‧second terminal voltage

Cse‧‧‧ series capacitor

Cp‧‧‧Shut capacitor

Csp‧‧‧Short shunt capacitor

Clp‧‧‧ long parallel capacitor

I‧‧‧ individuals

N‧‧‧Quantity

Is‧‧‧ Selected Individuals

Iu‧‧‧Unsuccessful individuals

Iss‧‧‧ Selected Individuals

Ims‧‧‧ Mutant Individual

Ies‧‧‧Selected individuals

Ic‧‧‧ close to the individual

Rec‧‧‧ error convergence range

INmax‧‧‧Maximal iterations

VV‧‧‧Vacuable rate of change

TVV‧‧‧ overall voltage change rate

f(k)‧‧‧ fitness

(S10)~(S70)‧‧‧ steps

1 is a circuit diagram of a single-phase induction generator according to a first embodiment of the present invention; FIG. 2 is a flow chart of a self-excited capacitor configuration method for a single-phase induction generator according to a first embodiment of the present invention; FIG. 4 is a flow chart of a method for configuring a self-excited capacitor of a single-phase induction generator according to a second embodiment of the present invention; FIG. 5 is a third embodiment of the present invention; The single-phase induction generator circuit structure diagram; FIG. 6 is a flow chart of the self-excited capacitance configuration method of the single-phase induction generator according to the third embodiment of the present invention.

The invention proposes the optimal self-excited capacitance based on the Enhanced Genetic Algorithm for the self-excited capacitance candidate values of the self-excited connection modes of the three single-phase induction generators (Single-Phase Induction Generator). The configuration method is used to improve the voltage change rate of the self-excited single-phase induction generator to obtain the best operating efficiency. The three wiring methods are (1) double-winding self-excitation-main winding series capacitor architecture, (2) double-winding self-excitation-main winding short-parallel architecture, and (3) dual-winding self-excitation-main winding long parallel architecture.

The optimal self-excited capacitor configuration method proposed by the invention can reduce the overall voltage change rate of the load end of the self-excited single-phase induction generator under the limitation of the operating conditions of the variable-speed and variable load of the single-phase induction generator. lowest. The capacitor optimization strategy will be determined by an enhanced gene algorithm. The optimum value of the shunt capacitance of the auxiliary winding is the optimum value of the series capacitor and the shunt capacitor of the main winding as the configuration of the self-excited capacitor.

Gene algorithm is an adaptive, heuristic, random, comprehensive based on Darwin's "Natural Evolution" and Mendes's "Genetic Variation" theory. Optimize the search algorithm. The gene algorithm has the characteristics of high efficiency, parallel processing and comprehensive search through mechanisms such as evaluation, selection, mating and mutation. In the gene algorithm, each individual is called a chromosome (Chromosome), and the gene value of each chromosome is generated by random numbers, and the collection of all chromosomes of each generation is called a population. In each generation, each chromosome competes with each other, and the chromosomes that are more suitable for survival have a higher fitness value. The chromosomes with higher fitness have the right to copy more sub-generations and then select two from them. Chromosome pairing for crossover produces the next generation of individuals, with a view to producing a more adaptive next generation of individuals. In order to avoid missing certain useful features, the treatment of Mutation is added to produce individuals with other characteristics. The evolution of such a generation will eventually lead to the best fitness for the individual, which is the best solution for genetic algorithm search.

The Enhanced Genetic Algorithm (EGA) is an enhanced version of the gene algorithm with adaptive mating and mutation. The enhanced gene algorithm adds adaptive function in the process of searching for the best individual. The enhanced gene algorithm adaptively adjusts the probability of mating and mutation according to different conditions of the individual to maintain the diversity of the group and prevent premature convergence. Further, the operation speed and accuracy of the work can be enhanced. The optimal solution search for enhanced gene algorithms can be divided into the following five mechanisms, which are briefly described as follows:

A. Initial group

The initial population is a collection of potentially feasible solutions. The initial set of populations is randomly generated. The set of potential feasible solutions contains some candidate chromosomes called individuals, and each individual's parameters have several genes.

B. Evaluation

Each individual in the population will be evaluated by a fitness function, and the present invention will assess the fitness of each individual in the population with the complement of the Total Voltage Variation (TVV) of the single phase induction generator.

C. Select

Based on the higher fitness of the father will produce a better theoretical basis of the offspring, the purpose of the selection mechanism is to select excellent individuals in the current group, and the outstanding individuals will have the opportunity to be the descendants of the next generation. Through the selection mechanism, the enhanced gene algorithm shows the Darwin's theoretical principles of superiority and defeat, and individuals with high fitness will be selected. After the selection process, individuals with high fitness will perform a mating mechanism to generate a new generation of individuals from a high-compatibility father.

D. Mating

The mating mechanism is the most important genetic operation in the enhanced gene algorithm. The mating mechanism exchanges two different individuals at the same selected position and then generates a new individual. Through the mating mechanism, a new generation of individuals combines the individual characteristics of the two parents, and the probability of mating is determined according to (1):

Where fmax is the maximum fitness of each individual in the population, favg is the average fitness of each individual body, and f is the greater fitness for mating two individuals. In practical applications, usually K1=K3=1, the value of Pc Usually in the range of 0.5-1.0.

E. Mutation

The mutation mechanism provides an opportunity to generate other individual characteristics in a new generation. Several individuals are randomly selected in the mutation mechanism, and then the selected individual is genetically altered at a specific location to generate a new individual. The probability of mutation is based on (2 ) decision:

Where f is the fitness of individual mutations. In practical applications, usually K2 = K4 = 0.05, and the value of Pm is usually in the range of 0.005-0.05.

After selection, mating, and mutation, a new generation of offspring will be created. This new generation will repeat the same selection, mating, and mutation processes. This iterative process will have reached the maximum number of iterations or When the error value convergence criterion has been entered, it is terminated.

The invention proposes a set of optimal self-excited capacitor configuration method based on the enhanced gene algorithm, and the following is directed to (1) the best auxiliary winding parallel capacitor of the double-winding self-excitation-main winding series capacitor structure is connected in series with the main winding. Capacitor configuration method, (2) single-winding self-excitation-main winding short-parallel architecture, optimal auxiliary winding shunt capacitor and main winding short-parallel series capacitor and shunt capacitor configuration method, (3) single-winding self-excitation-main winding long parallel architecture The optimal auxiliary winding shunt capacitor and the main winding long parallel series capacitor and shunt capacitor configuration method are detailed. The goal of optimizing the self-excited capacitor configuration is to minimize the voltage variation to improve the operating efficiency of the single-phase induction generator. .

1 is a circuit diagram of a single-phase induction generator according to a first embodiment of the present invention. As shown in FIG. 1, the single-phase induction generator 100 is a two-winding self-excited-main winding series capacitor structure. The single-phase induction generator 100 of the main winding series capacitor includes a rotor 10 and a stator 20, and the rotor 10 is driven by a prime mover 30. The prime mover 30 drives the rotor 10 in the direction of the magnetic field rotation and causes the induction motor to enter the induction generator 100 when its rotational speed exceeds the synchronous rotational speed. The stator 20 includes a main winding 21 and an auxiliary winding 22. The main winding 21 is connected in series with a series capacitor Cse and then externally connected to a load 23. The auxiliary winding 22 is connected in parallel with a parallel capacitor Cp. As shown in Figure 1, the single phase induction generator 100 employs a dual winding architecture. When the rotor 10 rotates, a terminal voltage Vo is generated across the load 23. Therefore, the terminal voltage Vo generated at both ends of the load 23 is determined according to the formula (3): V o = I _L × ( R _L + jX _L ) (3)

Among them, the IL in the formula (3) is the load current, RL is the load resistance, and XL is the load reactance. The method for obtaining the load current IL in the formula (3) and the equivalent circuit of the series winding capacitor Cse of the main winding 21 and the impedance algorithm of the equivalent circuit can be obtained by a conventional two-symmetric component equivalent circuit or a dq-axis equivalent circuit. (d-axis & q-axis equivalent circuit) is found and will not be described here.

The object of the present invention is to find the shunt capacitance Cp and the series capacitor Cse which can maintain the minimum voltage change rate TVV of the load end of the single-phase induction generator 100 of the main winding series capacitor in the case of varying the rotational speed and the variable load. Therefore, firstly, the voltage change rate (VV) of the single-phase induction generator 100 of the main winding series capacitor is first obtained, and the calculation formula of the voltage change rate is as follows: VV = (| V o|-1) ² (4 )

The range of the rotational speed of the conventional induction generator 100 is usually between synchronous speed and 1.2 times synchronous speed. Therefore, taking the synchronous speed to 1.2 times the synchronous speed as an example, the speed v(j) can be set by the following formula (resolution is 1000):

Where v is the synchronous speed, v(j) is the jth speed, j=1, 2, . . . , 1000. It is worth mentioning that the range of the rotation speed of the induction generator 100 is not limited to the synchronous rotation speed to 1.2 times the synchronous rotation speed, and is merely taken as an example for convenience of explanation. In other words, as long as the single-phase induction motor can achieve the speed of power generation, it should be included in the scope of this embodiment. In addition, the resolution is set to 1000, and the representative rotation speed is divided into 1000 equal parts from the synchronous rotation speed to the 1.2 times synchronous rotation speed, but is not limited thereto, and is merely an example for convenience of explanation.

The variation of the load 23 of the conventional induction generator 100 is usually between the rated load and the rated load of 0.01 times. The power factor is cos θ as an example. The load resistor RL(i) and the load reactance XL(i) can be set by the following equation (resolution is 1000):

Where ZL is the rated load impedance, RL(i) is the ith load resistance, and XL(i) is the ith load reactance, i=1, 2, . . . , 1000. It is worth mentioning that the variation range of the load 23 of the induction generator 100 is not limited to the rated load to 0.01 times the rated load, and is merely taken as an example for convenience of explanation. In addition, the resolution is set to 1000, and the representative speed is divided into 1000 equal parts from the rated load to 0.01 times the rated load, but is not limited thereto, and is merely an example for convenience of explanation.

The objective of applying the enhanced gene algorithm of the present invention is to maintain the overall voltage change rate TVV minimum in the case of varying rotational speed and variable load, and in this embodiment, the rotational speed range is set to synchronous rotational speed to 1.2 times synchronous rotational speed, load 23 range setting. For rated load to 0.01 times rated load, the overall voltage change rate TVV can be expressed as:

It is worth mentioning that the magnitude of the above resolution represents the length of time required to calculate the above parameters and the accuracy of the overall voltage change rate TVV obtained. Therefore, if the resolution is low, the time calculated by the representative parameter is short, but the accuracy of determining the overall voltage change rate TVV is low. If the resolution is high, the time calculated by the representative parameter is long, but the accuracy of the overall voltage change rate TVV is high.

2 is a flow chart of a method for configuring a self-excited capacitor of a single-phase induction generator according to a first embodiment of the present invention. Referring to FIG. 1, the present invention finds the optimal solution of the parallel capacitance Cp of the auxiliary winding 22 and the series capacitance Cse of the main winding 21 with the optimization characteristics of the enhanced gene algorithm. The basic concept is that The parallel capacitance Cp of the auxiliary winding 22 and the series capacitance Cse of the main winding 21 are set as parameters of each individual I of the enhanced gene algorithm, and the optimal solution is found through the optimization process of the group to determine the parallel capacitance Cp of the auxiliary winding 22 and The main winding 21 is connected to the final value of the capacitor Cse. The optimization process includes setting the number of groups, defining individual parameters, setting initial values, calculating, selecting, mating, and abrupt changes in fitness functions. The method flow can be implemented by an electronic device (not shown) including an executable, select, calculate, and determine function, and the obtained result can be applied to the three types of induction generators 100.

As shown in Figure 2, and with reference to Figure 1. The self-excited capacitor configuration method includes: first, setting a plurality of individuals, each of which includes a candidate capacitance of the series capacitor and a candidate capacitance of the parallel capacitor (S10). The number N of the individuals I represents the number N of the individuals I participating in the competition during each iteration, and the number of the numbers N directly affects the speed of the solution. The more the number N, the longer the convergence time, but it is easy to obtain the overall optimal solution. Conversely, if the number N is set too small, it can reduce the convergence time, but it is easy to fall into the local optimal solution. In order to avoid the number N of the individual I, the convergence time is prolonged, and the number N of the individuals I is too small to fall into a local optimum solution. Therefore, in the present embodiment, the number N of the individuals I is set to 100. Each individual I contains a candidate capacitance of the parallel capacitance Cp of the auxiliary winding 22 and a candidate capacitance of the series capacitance Cse of the main winding 21. Then, the candidate capacitance of the series capacitance of the individual and the candidate capacitance of the parallel capacitance are determined in a random manner (S20). The self-excited capacitor configuration method randomly configures the candidate capacitance of the parallel capacitance Cp of the auxiliary winding 22 and the candidate capacitance of the series capacitance Cse of the main winding 21 in a random number manner. And calculating a first terminal voltage Vo1 of the load 23 of each individual I by substituting the candidate capacitance of the parallel capacitance Cp of each individual I and the candidate capacitance of the series capacitance Cse into the above formula (3). It is worth mentioning that the range of existing capacitor candidate values is mostly between picofara (pF) and millifarad (milli F; mF). However, in order to avoid the parameter range determined by the random number being too large, the candidate capacitance of the parallel capacitance Cp and the candidate capacitance of the series capacitance Cse between each individual I are too extreme, so the random number determines the capacitor candidate capacitance range to 100μF to 200μF is optimal.

Refer to Figure 2 and refer to Figure 1. Then, a first terminal voltage of the load of each individual is calculated by the candidate capacitance of the series capacitance of each individual and the candidate capacitance of the parallel capacitance, and each individual is calculated by the first terminal voltage Overall voltage change rate (S30). The voltage change rate VV of each individual I can be obtained by substituting the first terminal voltage Vo1 of the load 23 of each individual I into the above equation (4). Then, the voltage change rate VV of each individual I is represented by the above formula (8), and the overall voltage change rate TVV of each individual I in the case of the varying rotational speed and the varying load is obtained. By the enhanced gene algorithm, the fitness fV of each individual I can be calculated from the overall voltage change rate TVV of each individual I. In this embodiment, the complement of the overall voltage change rate TVV of each individual I is the fitness f(k), and the calculation of f(k) is as follows: f ( k )=1- TVV ( k ) for k =1,2,......., N (9)

Where k is the number of the kth individual, N individual I. In general, the smaller the value of the overall voltage change rate TVV, the smaller the variation of the overall voltage. Therefore, the smaller the value of the overall voltage change rate TVV, the better. As can be seen from the above formula (9), the larger the value of the fitness f(k), the smaller the fluctuation range of the overall voltage. Therefore, the larger the value of the fitness f(k), the better.

Please refer to Figure 2 and refer to Figure 1. After the above steps, it can be known that each body I obtains the candidate capacitance of each parallel capacitance Cp and the candidate capacitance of the series capacitance Cse, and the overall voltage change rate TVV of each body I. Then, a plurality of selected individuals are selected by the overall voltage change rate of each individual, and the unselected individuals are a plurality of unsuccessful individuals (S40). After the overall voltage change rate TVV of each body I is obtained, the overall voltage change rate TVV of each individual I is sorted from large to small, and the overall voltage change rate TVV of the individual I is smaller, representing the parallel capacitance in the individual I. The candidate capacitance of Cp and the candidate capacitance of series capacitor Cse are better, so there is a greater probability of selection. When a plurality of selected individuals Is are selected, the remaining unselected individuals I are classified into a plurality of unsuccessful individuals Iu. In the present embodiment, 20% of the selected individuals Is are selected, and the remaining 80% are the selected individuals Iu.

After the above selection steps, the system enters the stage of mating and mutation. Then, the candidate capacitance of the series capacitance or the candidate capacitance of the parallel capacitance of each selected individual is randomly exchanged, and a plurality of selected sub-individuals are generated (S50). After selecting the selected individuals Is, the candidate capacitances of the parallel capacitors Cp in the selected individuals Is or the candidate capacitances of the series capacitors Cse in the selected individuals Is are randomly exchanged. The mating probability is determined by the above formula (1). Therefore, the overall voltage change rate TVV of the selected individuals Is needs to be converted into the fitness f(k), and then the above formula (1) is brought to obtain the mating probability. As can be seen from the above formula (1), the lower the overall voltage change rate TVV, the higher the mating probability. After mating, a new individual is generated as a plurality of selected sub-individual Iss, and the selected sub-individual Iss retains the candidate capacitance of the parallel capacitance Cp of one of the selected individuals Is or the candidate capacitance of the series capacitance Cse . Then, at least one of the selected individuals is randomly selected as at least one mutant individual, and the candidate capacitance of the series capacitance or the candidate capacitance of the parallel capacitance (S50') in the individual of the mutants is randomly changed. Although the missing individuals Iu may represent the overall voltage change rate TVV is not good, but the reason why the overall voltage change rate TVV is low is due to the candidate capacitance or series capacitance Cse of the parallel capacitance Cp of the selected individuals Iu. The influence of one of the candidate values. Therefore, the mechanism for retaining the mutation is to try to change whether the organic potential can mutate the candidate capacitance of the parallel capacitor Cp and the series capacitance Cse by changing one of the candidate capacitance of the parallel capacitance Cp or the candidate capacitance of the series capacitance Cse. Candidate values are a combination of preferred individuals. Therefore, among the above-mentioned selected individuals Iu, at least one of the missing individuals Iu is randomly selected, and one of the candidate capacitances of the parallel capacitance Cp or the candidate capacitance of the series capacitance Cse is changed to be at least one mutant individual Ims. The mutation probability is determined by the above formula (2). Therefore, the overall voltage change rate TVV of the selected individuals Iu needs to be converted into the fitness f(k), and the mutation probability is obtained by introducing the above formula (2). As can be seen from the above formula (2), the lower the overall voltage change rate TVV, the higher the probability of mutation. It is worth mentioning that comparing the above formula (1) (2), the probability of mating is much greater than the probability of mutation. Therefore, the number of selected individual individuals Iss is usually also greater than the number of Ims of the mutant individual.

Refer to Figure 2 and refer to Figure 1. Then, a second terminal voltage of the load of each of the selected sub-individuals and each of the mutant sub-individuals is calculated by the step (S30), and is calculated by the second terminal voltage The overall voltage change rate of each of the selected sub-individuals and each of the mutant sub-populations (S60). After the selected sub-individual Iss and the mutant sub-individual Ims are taken out from the stage of mating and mutation, the calculation method of step (S30) is used to obtain the Iss of each of the selected sub-individuals and the mutant sub-individual Ims. A second terminal voltage Vo2 of the load 23. Then, by the calculation method of the step (S30), and by the second terminal voltage Vo2, the overall voltage change rate TVV of each of the selected sub-individual Iss and the mutant sub-instant is calculated.

After the above-mentioned mating and mutation stages, and the overall voltage change rate TVV is obtained, the determination step is entered. Finally, sorting the selected individuals, the selected sub-individuals, and the magnitude of the overall voltage change rate of the individual mutants, and selecting a selected individual whose overall voltage change rate is the smallest, and determining the voltage of the selected individual Whether the rate of change is within an error convergence range (S70). By the above steps (S30) to (S60), the overall voltage change rate TVV of each of the selected sub-individual Iss, the mutant sub-indivis, and the selected individual Is can be obtained. Thereafter, the size of the overall voltage change rate TVV is sorted, and a selected individual Ies having the smallest overall voltage change rate TVV is taken out, and it is determined whether the overall voltage change rate TVV of the selected individual Ies is within an error convergence range Rec (error) Within the convergence range). Among them, the error convergence range Rec can set the following aspects. (1) It is determined whether the overall voltage change rate TVV of the individual Ies is less than 10%. (2) Compare whether the difference between the overall voltage change rate TVV of the individual Ies determined by the selected individual Ies and the next process is less than 10%. It is worth mentioning that the error convergence range Rec is not limited to the above. In other words, the operator can self-adjust the proportion of the error convergence range Rec. For example, but not limited to, the operator adjusts the error convergence range Rec to 3% to obtain a better candidate capacitance of the parallel capacitance Cp and a candidate capacitance of the series capacitance Cse, so as to more accurately control the induction generator 100. It is worth mentioning that the number of individual I in this embodiment is set to 100. Therefore, the total number of the selected sub-individual Iss, the mutant sub-interjects Ims, and the selected individual Iss obtained by the above steps is equivalent to the number of the individual I being optimal.

After the stage of the above judgment, if it is determined that the voltage change rate TVV of the individual Ies is within the error convergence range Rec, the candidate capacitance and the series capacitance representing the parallel capacitance Cp of the selected individual Ies are determined. The candidate value of Cse is the best solution. At this time, the candidate capacitance of the series capacitor Cse representing the selected individual Ies and the candidate capacitance of the parallel capacitor Cp are selected as the configuration of the self-excited capacitor. If it is determined that the voltage change rate TVV of the individual Ies is not within the error convergence range Rec, the selected individuals Is obtained in the steps (SI0) to (S70), the selected sub-individual Iss, and The mutant individual Ims is set to the individuals I, and jumps back to the step (S30). After repeating the steps (S30) to (S70), it is determined that the voltage change rate TVV of the selected individual Ies selected in the one-step repeating step (S30) to the step (S70) is within the error convergence range Rec; the flow (S30) The process of (S70) is repeated to determine that the voltage change rate TVV of the selected individual Ies is within the error convergence range Rec. If the step (S30) to the step (S70) are repeatedly performed over a maximum number of iterations INmax, it may represent that the candidate capacitance of the series capacitor Cse in the individual I may be randomly determined in steps (S10) to (S20). The candidate capacitance of the parallel capacitor Cp is inferior, resulting in the subsequent mating and mutation steps, and the selected sub-individual Iss and the mutants are not good. Therefore, among the selected individuals Is, the selected individual individuals Iss, and the mutant individuals Im, the closest individual Ic closest to the error convergence range Rec is selected. At this time, the candidate capacitance of the series capacitor Cse representing the proximity individual Ic and the candidate capacitance of the parallel capacitance Cp may be selected as the configuration of the self-excited capacitor; or a plurality of individuals I may be reselected, and the step (S10) is performed again. To (S70), an attempt is made to find the confirmed individual Ies within the error convergence range Rec. It is worth mentioning that in order to avoid the repetition of the steps (S30) to (S70), the time is too long or redundant; or the number of times of repeating the steps (S30) to (S70) is insufficient, which results in the error of the selected individual Ies. Convergence range within Rec. Therefore, in the present embodiment, it is preferable that the maximum number of iterations INmax is set to 2000 times.

Please refer to FIG. 3, which is a circuit diagram of a single-phase induction generator according to a second embodiment of the present invention. As shown in FIG. 3, the single-phase induction generator 100 is a two-winding self-excitation-main winding short shunt capacitor architecture. The short-parallel capacitor single-phase induction generator 100A includes a rotor 10A and a stator 20A, and the rotor 10A is driven by a prime mover 30A. The prime mover 30A drives the rotor 10A in the direction of the magnetic field rotation, and when the rotational speed exceeds the synchronous rotational speed, the induction motor enters to operate the induction generator 100A. The stator 20A includes a main winding 21A and an auxiliary winding 22A. The main winding 21A is connected in parallel with a short parallel capacitor Csp, and then a series capacitor Cse is connected in series with an external load 23A. The auxiliary winding 22A is connected in parallel with a parallel capacitor Cp. As shown in FIG. 3, the single phase induction generator 100A employs a dual winding architecture. When the rotor 10A rotates, a terminal voltage Vo is generated across the load 23A. The difference between this embodiment and the first embodiment (please refer to FIG. 1) is that a short parallel capacitor Csp is connected in parallel in the main winding 21A, but the load 23A end can be reduced compared with the architecture of the first embodiment. The voltage Vo is affected by the load current. In the present embodiment, the algorithm and the steps of determining the overall voltage change rate TVV of the load 23A by the terminal voltage Vo of the load 23A are the same as those in the first embodiment, and will not be further described herein.

4 is a flow chart of a method for configuring a self-excited capacitor of a single-phase induction generator according to a second embodiment of the present invention. Referring to FIG. 3, the present invention finds the optimal solution of the parallel capacitance Cp of the auxiliary winding 22A, the series capacitance Cse of the main winding 21A and the short parallel capacitance Csp by the optimization function of the genetic algorithm. The basic concept is to use the auxiliary winding 22A. The parallel capacitor Cp, the main winding 21A series capacitor Cse and the short parallel capacitor Csp are set to the parameters of each individual I of the enhanced gene algorithm, and the optimal solution is found through the group optimization process to determine the parallel capacitance of the auxiliary winding 22A. Cp, the final value of the series winding Cse of the main winding 21A and the short parallel capacitor Csp. The optimization process includes setting the number of groups, defining individual parameters, setting initial values, calculating, selecting, mating, and abrupt changes in fitness functions.

As shown in FIG. 4, and referring to FIG. 3, the self-excited capacitor configuration method includes: first, setting a plurality of individuals, each of which includes a candidate capacitance value of the short winding capacitance of the main winding, and a candidate capacitance value of the series capacitor. A candidate capacitance value with the shunt capacitor (S10). Referring to FIG. 4 and referring to FIG. 2 to FIG. 3, the maximum difference between the self-excited capacitor configuration method of this embodiment and the first embodiment (see FIG. 2) is that the main windings 21A are short-paralleled in the individual I. The candidate capacitance of the capacitor Csp. The flow of the step (S10) to the step (S70) and the judgment manner are the same as those of FIG. 2 except that the parameters of the candidate capacitance of the short-parallel capacitor Csp of the main winding 21A are different, and details are not described herein again.

Please refer to FIG. 5, which is a circuit diagram of a single-phase induction generator according to a third embodiment of the present invention. As shown in FIG. 5, the single-phase induction generator 100 is a two-winding self-excitation-main winding long parallel capacitor structure. The single-phase induction generator 100B of the long parallel capacitor includes a rotor 10B and a stator 20B, and the rotor 10B is driven by a prime mover 30B. The prime mover 30B drives the rotor 10B in the direction of the magnetic field rotation, and when the rotational speed exceeds the synchronous rotational speed, the induction motor enters the induction generator 100B. The stator 20B includes a main winding 21B and an auxiliary winding 22B. The main winding 21B is connected in series with a series capacitor Cse, and then a long parallel capacitor Clp is connected in parallel with an external load 23B. The auxiliary winding 22B is connected in parallel with a parallel capacitor Cp. As shown in FIG. 5, the single phase induction generator 100B employs a dual winding architecture. When the rotor 10B rotates, a terminal voltage Vo is generated across the load 23B. The difference between this embodiment and the first embodiment (please refer to FIG. 1) is that a long parallel capacitor Clp is connected in parallel with the main winding 21B, but the load 23B can be reduced compared with the architecture of the first embodiment. The voltage Vo is affected by the load current. In the present embodiment, the algorithm and the steps of determining the overall voltage change rate TVV of the load 23B by the terminal voltage Vo of the load 23B are the same as those in the first embodiment, and will not be further described herein.

6 is a flow chart of a method for configuring a self-excited capacitor of a single-phase induction generator according to a third embodiment of the present invention. Referring to FIG. 5, the present invention finds the parallel connection of the auxiliary winding 22B by using the optimization function of the genetic algorithm. The optimal solution of capacitor Cp, main winding 21B series capacitor Cse and long parallel capacitor Clp, the basic concept is to set the parallel capacitor Cp of the auxiliary winding 22B, the main winding 21B series capacitor Cse and the long parallel capacitor Clp as the enhanced gene algorithm. The parameters of each individual I find the optimal solution through the optimization process of the group to determine the final value of the parallel capacitance Cp of the auxiliary winding 22B, the series capacitance Cse of the main winding 21B and the long parallel capacitance Clp. The optimization process includes setting the number of groups, defining individual parameters, setting initial values, calculating, selecting, mating, and abrupt changes in fitness functions.

As shown in FIG. 6 and referring to FIG. 5, the self-excited capacitor configuration method includes: first, setting a plurality of individuals, each of which includes a candidate capacitance of the main winding long parallel capacitor, and a candidate capacitance of the series capacitor. A candidate capacitance value with the shunt capacitor (S10). Please refer to Figure 6, and with reference to Figures 2 and 5, this The biggest difference between the self-excited capacitor configuration method of the embodiment and the first embodiment (please refer to FIG. 2) is that the candidate capacitance of the long parallel capacitor Clp of the main winding 21B is increased in the individual I. The process of the steps (S10) to (S70) and the determination manner are the same as those of FIG. 2 except that the parameters of the candidate capacitance of the long parallel capacitance Clp of the main winding 21B are different, and details are not described herein again.

In summary, the present invention has the following advantages: 1. Compared with the current attempt to determine the candidate capacitance of the self-excited capacitor, the self-excited capacitor configuration method provided by the present invention does not need to repeatedly replace the capacitance value and test the overall voltage. The step of varying the rate of TVV, therefore, the efficiency of the candidate capacitance of the series capacitor and the candidate capacitance of the parallel capacitor can be quickly obtained; 2. The present invention is compared to the current attempt to determine the candidate capacitance of the self-excited capacitor. The self-excited capacitor configuration method provided can gradually approach the optimal overall voltage change rate TVV, so that the candidate capacitance value of the series capacitor and the candidate capacitance value of the parallel capacitor can be accurately obtained. 3. The enhanced type is utilized. Due to the mutation step of the algorithm, the candidate capacitance of the series capacitor and the candidate capacitance of the parallel capacitor are prevented from falling into the local optimal solution. 4. The candidate capacitance of the precision series capacitor is used in parallel with the parallel capacitance. The candidate capacitance of the capacitor can improve the overall power generation efficiency of the induction generator 100.

However, the above description is only for the detailed description and the drawings of the preferred embodiments of the present invention, and the present invention is not limited thereto, and is not intended to limit the present invention. The scope of the patent application is intended to be included in the scope of the present invention, and any one skilled in the art can readily appreciate it in the field of the present invention. Variations or modifications may be covered by the patents in this case below.

Claims (12)

A self-excited capacitor configuration method for a single-phase induction generator, the single-phase induction generator includes a main winding and an auxiliary winding, the main winding is connected in series with a series capacitor and then connected in parallel with a load, the auxiliary winding is connected in parallel with a parallel capacitor The self-excited capacitor configuration method includes: (a) setting a plurality of individuals, each individual including a candidate capacitance of the series capacitor and a candidate capacitance of the parallel capacitor; (b) determining the individuals in a random manner a candidate capacitance of the series capacitor and a candidate capacitance of the parallel capacitor; (c) calculating a first of the load of each individual by the candidate capacitance of the series capacitance of each individual and the candidate capacitance of the parallel capacitance a terminal voltage, and calculating an overall voltage change rate of each individual by the first terminal voltage; (d) selecting a plurality of selected individuals by an overall voltage change rate of each individual, and the plurality of unselected individuals are plural (el) randomly selecting the candidate capacitance of the series capacitance of each selected individual or the candidate capacitance of the parallel capacitance, and generating a plurality of selected sub-objects; (e2) randomly selecting the selected individuals at least Each of the at least one mutant individual, and randomly changing the candidate capacitance of the series capacitance or the candidate capacitance of the parallel capacitance within the individual of the mutant; (f) calculating each of the selected sub-objects by step (c) And a second terminal voltage of the load of each of the mutant sub-populations, and calculating an overall voltage change rate of each of the selected sub-individents and each of the mutant sub-populations by the second terminal voltage; (g) sorting the selected individuals, the selected sub-individuals, and the magnitude of the overall voltage change rate of the individual mutants, and selecting a selected individual having the smallest overall rate of change of the voltage, and determining the individual of the selected individual Whether the voltage change rate is within an error convergence range; wherein, if the voltage change rate of the selected individual is within the error convergence range, selecting a candidate capacitance value of the series capacitor representing the selected individual and a candidate for the parallel capacitance The capacitance is used as a configuration of the self-excited capacitor. The method for configuring a self-excited capacitor according to claim 1, wherein the step (g) further comprises: (g1) selecting the selected individual if the voltage change rate of the selected individual is not within the error convergence range The selected individual individuals and the individual mutants are set as the individuals and jump back to step (c). The self-excited capacitor configuration method according to claim 2, wherein after the step (g1), the method further comprises: (g2) if the step (c) to the step (g1) are repeatedly performed over a maximum number of iterations, Among the selected individuals, the selected sub-individuals, and the individual individuals of the mutants, the closest individual to the error convergence range, and the candidate capacitance of the series capacitance representing the proximity individual and the candidate of the parallel capacitance are selected. Capacitance. The self-excited capacitor configuration method according to claim 1, wherein the step (d) further comprises: (d1) ordering the overall voltage change rate of each individual from large to small, and the overall voltage change rate of the individuals. The smaller, the greater the chance of selection. A self-excited capacitor configuration method for a single-phase induction generator, the single-phase induction generator includes a main winding and an auxiliary winding, the main winding is connected in parallel with a main winding short shunt capacitor and then connected in series with a series capacitor and a load, The auxiliary winding is connected in parallel with a parallel capacitor, and the self-excited capacitor configuration method includes: (a) setting a plurality of individuals, each individual comprising a candidate capacitance of the short winding capacitance of the main winding, a candidate capacitance of the series capacitance and a candidate capacitance of the parallel capacitance; (b) determining the individuals in a random manner a candidate capacitance of the short-parallel capacitor of the main winding, a candidate capacitance of the series capacitor and a candidate capacitance of the parallel capacitor; (c) a candidate capacitance of the short-parallel capacitor of the main winding of each individual, the series capacitor a candidate capacitance value and a candidate capacitance value of the parallel capacitance calculate a first terminal voltage of the load of each individual, and calculate an overall voltage change rate of each individual by the first terminal voltage; (d) by each The individual's overall voltage change rate selects a plurality of selected individuals, and the unselected individuals are a plurality of unsuccessful individuals; (e1) randomly exchanges the candidate capacitance of the short winding capacitance of the main winding of each selected individual, the series a candidate capacitance of the capacitance or a candidate capacitance of the parallel capacitance, and generating a plurality of selected sub-subjects; (e2) randomly selecting at least one of the unselected individuals as at least one mutant sub-person, and randomly changing the mutants In vivo a candidate capacitance of the short-parallel capacitance of the main winding, a candidate capacitance of the series capacitance or a candidate capacitance of the parallel capacitance; (f) calculating, by step (c), each of the selected sub-individuals and each of the mutant sub-individuals a second terminal voltage of the load, and calculating, by the second terminal voltage, an overall voltage change rate of each of the selected sub-individuals and each of the mutant sub-populations; (g) sorting the selected individuals, the selected ones The magnitude of the overall voltage change rate of the individual and the individual mutants, and selecting a selected individual whose overall voltage change rate is the smallest, and determining whether the voltage change rate of the selected individual is within a range of error convergence; When it is determined that the voltage change rate of the individual is within the error convergence range, the candidate capacitance value of the short-parallel capacitor of the main winding representing the selected individual, the candidate capacitance of the series capacitor and the candidate capacitance of the parallel capacitor are selected to As a configuration of self-excited capacitors. The self-excited capacitor configuration method according to claim 5, wherein the step (g) further comprises: (g1) if the voltage change rate of the selected individual is not within the error convergence range, the selected individuals are selected The selected individual individuals and the individual mutants are set as the individuals and jump back to step (c). The self-excited capacitor configuration method according to claim 6, wherein the step (g1) further comprises: (g2) if the step (c) to the step (g1) are repeatedly performed over a maximum number of iterations, selecting the Among the selected individuals, the selected sub-individuals, and the individual individuals of the mutants, the closest individual to the error convergence range is selected, and the candidate capacitance value of the short-parallel capacitance of the main winding representing the approaching individual is selected, and the series is selected. The candidate capacitance of the capacitor and the candidate capacitance of the shunt capacitor. The self-excited capacitor configuration method according to claim 5, wherein the step (d) further comprises: (d1) ordering the overall voltage change rate of each individual from large to small, and the overall voltage change rate of the individuals The smaller, the greater the chance of selection. A self-excited capacitor configuration method for a single-phase induction generator, the single-phase induction generator includes a main winding and an auxiliary winding, the main winding is connected in series with a series capacitor and then connected to a main winding long parallel capacitor and a load, The auxiliary winding is connected in parallel with a parallel capacitor. The self-excited capacitor configuration method comprises: (a) setting a plurality of individuals, each individual comprising a candidate capacitance of the series capacitor, a candidate capacitance of the main winding long parallel capacitor, and the parallel connection. a candidate capacitance of the capacitor; (b) determining, in a random manner, a candidate capacitance of the series capacitance of the individual, a candidate capacitance of the long parallel capacitance of the main winding, and a candidate capacitance of the parallel capacitance; (c) by each The candidate capacitance of the series capacitor, the candidate capacitance of the main winding long parallel capacitor, and the candidate capacitance of the parallel capacitor calculate a first terminal voltage of the load of each individual, and by the first end The voltage calculates the overall voltage change rate of each individual; (d) selecting a plurality of selected individuals by the overall voltage change rate of each individual, and the unselected individuals are a plurality of unsuccessful individuals; (e1) randomly exchanging candidate capacitance values of the series capacitance of each selected individual a candidate capacitance of the main winding long parallel capacitance or a candidate capacitance of the parallel capacitance, and generating a plurality of selected sub-individuals; (e2) randomly selecting at least one of the missing individuals as at least one mutant sub-individual, and Randomly changing the candidate capacitance of the series capacitance in the individual mutant sub-elements, the candidate capacitance of the main winding long parallel capacitance or the candidate capacitance of the parallel capacitance; (f) calculating each selection by step (c) a second terminal voltage of the load of the sub-individual and each of the mutant sub-individuals, and calculating an overall voltage change rate of each of the selected sub-individuals and each of the mutant sub-individuals by the second terminal voltage; (g) sorting the The size of the overall voltage change rate of the selected individuals, the selected sub-individuals, and the individual mutants, and selects a selected individual whose overall voltage change rate is the smallest, and determines whether the voltage change rate of the selected individual is Error convergence range Wherein, if the voltage change rate of the selected individual is within the error convergence range, the candidate capacitance of the series capacitor representing the selected individual, the candidate capacitance of the main winding long parallel capacitor, and the parallel capacitance are selected. The candidate capacitance is used as the configuration of the self-excited capacitor. The method for configuring a self-excited capacitor according to claim 9, wherein the step (g) further comprises: (g1) selecting the selected individual if the voltage change rate of the selected individual is not within the error convergence range The selected individual individuals and the individual mutants are set as the individuals and jump back to step (c). The self-excited capacitor configuration method according to claim 10, wherein the step (g1) further comprises: (g2) if the step (c) to the step (g1) are repeatedly performed over a maximum number of iterations, selecting the Among the selected individuals, the selected sub-individuals, and the individuals of the mutants, the one closest to the error convergence range Approaching the individual, and selecting a candidate capacitance representative of the series capacitance of the proximity individual, a candidate capacitance of the main winding long parallel capacitance, and a candidate capacitance of the parallel capacitance. The self-excited capacitor configuration method according to claim 9, wherein the step (d) further comprises: (d1) ordering the overall voltage change rate of each individual from large to small, and the overall voltage change rate of the individuals. The smaller, the greater the chance of selection.