Active harmonic suppression optimization method for water electrolysis hydrogen production rectifier transformer

Abstract

The invention discloses an active harmonic suppression optimization method for a water electrolysis hydrogen production rectifier transformer, and the method comprises the following steps: constructing a rectifier circuit which comprises a switching tube, combining a voltage vector and a zero vector outputted by the switching tube, and obtaining a combined output voltage vector; processing the combined output voltage vector to obtain a harmonic amplitude, introducing a Levy tangent flight strategy and a trust region variation strategy into a black-wing optimization algorithm to construct an improved dynamic search algorithm, and performing iterative calculation on the harmonic amplitude through the improved dynamic search algorithm to obtain a total harmonic distortion rate, namely an optimal switching tube combination. The switch tubes of the rectifier circuit are adjusted through the optimal switch tube combination, and the total harmonic distortion rate of the rectifier circuit is reduced. The optimal switch tube combination of the rectifier circuit is iteratively calculated through the improved dynamic search algorithm, resource consumption is effectively reduced, the processing speed is increased, and the efficiency and accuracy of the water electrolysis hydrogen production rectifier are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of harmonic of rectifier transformers, and particularly to an active harmonic suppression optimization method for rectifier transformers in water electrolysis hydrogen production. Background Art

[0002] As a secondary energy source, hydrogen energy has many advantages: it can not only be stored and used for long-distance transmission, but also be efficiently converted into electrical energy and heat energy; its energy density is very high; the final product of its reaction is water and no other pollutants are generated; in addition, hydrogen has a wide range of applications in the industrial field and is a key raw material.

[0003] Harmonic pollution not only seriously affects the power quality of the power grid, reduces the efficiency of power generation, transmission and utilization of electrical energy, but also harms electrical equipment, increases system losses, reduces the system power factor, accelerates the aging of insulating media such as cables and wires, and causes problems such as line-to-line short circuits, ground faults and misoperation of protection circuits.

[0004] Active harmonic suppression methods consider from the specific rectifier device itself, improve the rectifier topology structure, and reduce or eliminate the generation of harmonics. From a technical perspective, this method can be further divided into PWM rectification technology and multi-pulse rectification technology. The goal of PWM is to eliminate specific secondary harmonics, ensure that certain low-frequency harmonics are completely filtered out, but do not limit the total harmonic distortion (THD) and the amplitudes of other harmonics. All the constraint conditions are equality constraints, so the solution process is relatively simple. On the other hand, the latter requires that the amplitudes of all odd harmonics meet the grid standards, the constraint conditions include both equalities and inequalities, and pursues the lowest possible THD value, with the minimization of THD as the optimization goal. Therefore, studying how to perform harmonic suppression to improve its generalization ability, reduce resource consumption, and accelerate the processing speed plays an important role in improving the efficiency and accuracy of rectifiers for water electrolysis hydrogen production. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides an active harmonic suppression optimization method for a rectifier transformer in water electrolysis hydrogen production, which solves the problems mentioned in the background art.

[0006] To achieve the above object, the present invention provides the following technical solution: An active harmonic suppression optimization method for a rectifier transformer in water electrolysis hydrogen production, comprising the following steps: Step S1: Construct a rectifier circuit. The rectifier circuit includes switching tubes, combine the voltage vectors and zero vectors output by the switching tubes to obtain a combined output voltage vector; process the combined output voltage vector to obtain the harmonic amplitude; Step S2: Introduce the Levy tangent flight strategy and the trust region mutation strategy into the black-winged kite optimization algorithm to construct an improved dynamic search algorithm; Step S3: Iteratively calculate the harmonic amplitude through the improved dynamic search algorithm to obtain the total harmonic distortion rate, that is, the optimal switch tube combination. The switch tubes of the rectifier circuit are adjusted through the optimal switch tube combination to reduce the total harmonic distortion rate of the rectifier circuit.

[0007] Furthermore, the rectifier circuit includes an SVPWM circuit; the SVPWM circuit is divided into a plurality of sectors, each sector including a switch tube and a voltage space vector rotation angle, and the voltage vectors output by the switch tubes in the plurality of sectors are combined with a zero vector to obtain a combined output voltage vector; r is the voltage space vector rotation angle; is the combined output voltage vector; wherein the SVPWM circuit includes the midpoint voltage of the a-phase capacitor; is the point voltage.

[0008] Furthermore, based on the combined output voltage vector, let the switching period be T s ,express: (1); Where, ; represents the zero vector; represents the first zero vector; represents the second zero vector; Indicates the action time of the first zero vector U1; Represents the second zero vector duration of action; Represents the zero vector duration of action; Solve for the values of T1 and T2, which means: (2); (3); (4); Where, represents the modulation index; Indicates DC voltage; represents the sine function; Represents the cosine function.

[0009] Furthermore, the specific process of obtaining the harmonic amplitude is: In the process of calculating the harmonic amplitude of the SVPWM circuit, the Fourier decomposition of the point voltage is performed based on the combined output voltage vector, which is expressed as: (5); Where, Indicates point voltage over time The changing voltage waveform; represents the average value of the point voltage; represents infinity; represents the harmonic order; represents the fundamental angular frequency; and represent the amplitudes of the corresponding cosine term and sine term respectively; where ; is the fundamental frequency; The point voltage is an even function, which means: (6); (7); (8); In the formula, represents a small time increment; With the voltage space vector rotation angle as the boundary, the voltage waveform of the point voltage changing with time is expanded within half of the output switching period T and divided into Region 1 and s Region; Region; is the modulation factor; Based on the division into Region 1 and Region, the rotation angle of the voltage space vector corresponding to the kth region is set, which means: (9); In the formula, r k is the rotation angle of the voltage space vector corresponding to the kth region; represents the rotation angle increment of each interval; represents the total number of intervals; Let the positive level duty ratio of the rotation angle of the voltage space vector corresponding to the kth region be , which means: (10); In the formula, represents the starting moment of the time node of half of the switching period T s ; represents the last time node within half of the switching period T s , that is, the end moment of half of the period T s ; represents the end moment of the positive level duty ratio time node of the rotation angle of the voltage space vector corresponding to Region 1; represents the positive level duty ratio of the rotation angle of the voltage space vector corresponding to the 1st region; represents the end moment of the positive level duty ratio time node of the rotation angle of the voltage space vector corresponding to the 2k−2th region; Indicates the end moment of the positive-level duty cycle time node corresponding to the voltage space vector rotation angle in the (2k−3)-th sector; Indicates the positive-level duty cycle of the voltage space vector rotation angle corresponding to the (k−1)-th sector; Indicates the end moment of the positive-level duty cycle time node corresponding to the voltage space vector rotation angle in the (2k−1)-th sector; Indicates the period; Based on the positive-level duty cycle being and the period T, calculate and obtain the time value; Indicates the positive-level duty cycle of the first sector [[ID=1,6]]and the time value obtained from the period T, Indicates the positive-level duty cycle of the sector and the time value obtained from the period T; Then substitute the obtained time value into formula (8), expand the integral of the time value, and find , which represents: (11); In the formula, Indicates the amplitude of the n-th harmonic.

[0010] Furthermore, the specific process of introducing the Levy tangent flight strategy into the black-winged kite optimization algorithm is as follows: Set the parameters of the black-winged kite optimization algorithm. The parameters include the number of black-winged kites, the search space, the total number of iterations, and the leader. Generate a black-winged kite matrix based on the set parameters of the black-winged kite optimization algorithm. After randomly assigning the positions of each black-winged kite from the black-winged kite matrix, use the Tent map to generate the initial population, and use formula (12) to calculate the initial positions of the black-winged kites in the initial population, which represents: (12); In the formula, Indicates the initial position of the i-th black-winged kite; and respectively represent the lower and upper bounds of the search space; () represents a random number between [0, 1]; The optimal position y L of the black-winged kites in the initial population is calculated by the formula, which represents: (13); (14); In the formula, Indicates the currently found optimal fitness value; Indicates the fitness value of the i-th black-winged kite in the initial population; represents the black-winged kite; represents function; represents the minimum value; represents the fitness value of the i-th black-winged kite at the current position ; The attack strategy update equation is: (15); In the formula, represents the position of the i-th black-winged kite in the j-th dimension at the (t + 1)-th iteration step; represents the position of the i-th black-winged kite in the j-th dimension at the t-th iteration step; represents the threshold; represents a random number between 0 and 1 in the initial population; represents the non-linear factor; ; Q represents the total number of iterations; represents the base of the natural logarithm; represents otherwise; The black-winged kite migration strategy, which represents: (16); In the formula, represents a random number between 0 and 1, representing the Cauchy mutation; represents the one with the highest score among the black-winged kites in the j-th dimension at the t-th iteration step; represents the fitness value of the i-th black-winged kite; represents a randomly generated fitness threshold; represents a constant; ; The probability density function of the Cauchy mutation, which represents: (17); In the formula, represents the probability density function of the Cauchy distribution; represents the scale parameter; represents the random variable; represents the location parameter; When the probability density function becomes the standard form, which represents: (18); Adding the Levy tangent flight strategy algorithm to the black-winged kite migration strategy, the improved black-winged kite migration strategy, which represents: (19); In the formula, represents the tangent function; Indicates the flight length.

[0011] Furthermore, the specific process of introducing the prey escape strategy into the Black-winged Kite optimization algorithm is as follows: First, define the prey escape energy of the Black-winged Kite algorithm, which is expressed as: (20); In the formula, Indicates the prey escape energy; Indicates the exponential function; Indicates the ratio of the current iteration number t to the total iteration number ; When , it means that the prey does not have enough prey escape energy to escape the attack circle of the Black-winged Kite. At this time, the original attack strategy in formula (15) is used to update the position of the Black-winged Kite, where E0 represents the prey escape threshold; When , it means that the prey has enough prey escape energy to escape the attack of the Black-winged Kite. At this time, the trust region mutation strategy is introduced into the Black-winged Kite optimization algorithm to increase the search intensity and prevent the prey from escaping; First, define the Euclidean distance between Black-winged Kite i and Black-winged Kite , and calculate the mean position of all Black-winged Kites based on the Euclidean distance . According to the distance between the current Black-winged Kite i and the mean position , calculate the initial trust region radius, which is expressed as: (21); In the formula, Indicates the Euclidean distance between Black-winged Kite i and Black-winged Kite ; Indicates the Euclidean distance; Indicates the coordinate value of the i-th Black-winged Kite in the first dimension of the multi-dimensional space; Indicates the -th Black-winged Kite in the first dimension of the multi-dimensional space; Indicates the coordinate value of the i-th Black-winged Kite in the D-th dimension of the multi-dimensional space; Indicates the -th Black-winged Kite in the D-th dimension of the multi-dimensional space; Indicates the total number of Black-winged Kites; Indicates the positions of Black-winged Kite i and Black-winged Kite ; Indicates the initial trust region radius of the i-th Black-winged Kite; The distance between the current Black-winged Kite i and Black-winged Kite The Euclidean distance between them is less than the initial trust region radius of the \(i\)-th black-winged kite. As a trustworthy black-winged kite, calculate the mean position of the trustworthy black-winged kites, which is expressed as: (22); In the formula, represents the set of positions of the \(i\)-th trustworthy black-winged kite; represents the position of the \(i\)-th trustworthy black-winged kite; represents the mean position of the trustworthy black-winged kites; represents the number of trustworthy black-winged kites; represents the threshold of the mean position of the \(i\)-th trustworthy black-winged kite; Introduce the Levy tangent flight strategy in the attack strategy, which is expressed as: (23); (24); In the formula, represents the step size of the Levy tangent flight strategy algorithm; represents a random number that follows a normal distribution between [0, 1]; represents the standard deviation; represents the gamma function; represents a conventional parameter; Combine the Levy tangent flight strategy with the trust region mutation strategy to update the position of the attack strategy, which is expressed as: (25); The error of the position of the \(i\)-th black-winged kite in the \(j\)-th dimension at the \((t + 1)\)-th iteration after combination is corrected according to the optimal position \(y\) of the black-winged kites in the initial population L until the accuracy requirement of the black-winged kite optimization algorithm is met; When the black-winged kite optimization algorithm reaches the accuracy requirement, the iteration ends and the optimal identification parameters are obtained. Otherwise, return to formula (12) to recalculate the initial positions of the black-winged kites in the initial population. Considering that when the number of iterations of the black-winged kite optimization algorithm reaches the upper limit during the iteration process and the accuracy requirement is still not met, then select the minimum fitness value of the position of the \(i\)-th black-winged kite in the \(j\)-th dimension at the \((t + 1)\)-th iteration after combination during the iteration process as the optimal solution output.

[0012] Furthermore, the specific process of obtaining the total harmonic distortion rate is as follows: Iteratively calculate the \(n\)-th harmonic amplitude through the optimal identification parameters of the improved dynamic search algorithm , which is expressed as; (26); In the formula, represents the first harmonic amplitude; Indicates the total harmonic distortion rate of the first 50 harmonics of the output voltage; Indicates the error term of the harmonic amplitude; Indicates the pulse edge; Indicates the i-th switching device; Is the amplitude of the zero-th harmonic; Indicates the error term of the amplitude of the first harmonic; Z indicates the modulation ratio; Indicates voltage; Indicates the correction factor or scaling factor; Indicates the standard of the 5th harmonic amplitude; Indicates the error term of the 5th harmonic amplitude; Indicates the th harmonic amplitude standard; Indicates the th harmonic amplitude error term; = 1, 2, 3, …, 6; = 5, 7, 11, …, 49; Based on the total harmonic distortion rate of the first 50 harmonics of the output voltage Calculate the total harmonic distortion rate, which is expressed as: (27); In the formula, Indicates the total harmonic distortion rate; Indicates the absolute value of the harmonic amplitude.

[0013] Compared with the existing technology, the present invention has the following beneficial effects: (1), By adopting the Levy tangent flight strategy to simulate the flight behavior of the black-winged kite, the present invention can get rid of the dependence on the black-winged kite that has lost its leading role, which is beneficial to expanding the search range of the algorithm; introducing the prey escape strategy to optimize the black-winged kite optimization algorithm solves the problem that the black-winged kite optimization algorithm is prone to fall into a local optimal solution and cannot quickly converge to the optimal value in the later stage.

[0014] (2), The improved dynamic search algorithm of the present invention maintains a certain exploration ability in the early stage, thereby effectively reducing the total output harmonic distortion rate and showing strong adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 Is the method flow chart of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0016] As Figure 1 shown, the present invention provides a technical solution: A method for actively suppressing and optimizing harmonics of a rectifier transformer for hydrogen production by electrolyzing water, including the following steps: Step S1: Construct a rectifier circuit. The rectifier circuit includes a switching tube. Combine the voltage vectors and zero vectors output by the switching tube to obtain a combined output voltage vector. Process the combined output voltage vector to obtain the harmonic amplitude. Step S2: Introduce the Levy tangent flight strategy and the trust region mutation strategy into the black-winged kite optimization algorithm to construct an improved dynamic search algorithm. Step S3: Iteratively calculate the harmonic amplitude through the improved dynamic search algorithm to obtain the total harmonic distortion rate, that is, the optimal switching tube combination. Adjust the switching tubes of the rectifier circuit through the optimal switching tube combination to reduce the total harmonic distortion rate of the rectifier circuit.

[0017] Among them, the rectifier circuit includes an SVPWM circuit; use the space vector pulse width modulation SVPWM circuit in the active power filter APF as the research topology; divide the SVPWM circuit into 6 sectors. The 6 sectors include 6 switching tubes; the 6 sectors also include the voltage space vector rotation angle. Combine the voltage vectors and zero vectors output by the switching tubes in the 6 sectors to obtain a combined output voltage vector. Among them, the 6 switching tubes output 8 voltage vectors; 2 of the 8 voltage vectors are zero vectors, and the lengths of the 6 voltage vectors are 2 / 3Udc; r is the voltage space vector rotation angle. is the combined output voltage vector; among them, the SVPWM circuit includes the midpoint voltage of the a-phase capacitor. is the point voltage.

[0018] Among them, taking the combined output voltage vector as an example, let the switching period be T s , which means: (1); In the formula, ; represents the zero vector; represents the first zero vector; represents the second zero vector; represents the action time of the first zero vector U1; represents the second zero vector action time; represents the zero vector action time; Decompose formula (1) on the horizontal and vertical coordinates to obtain the values of T1 and T2, which means: (2); (3); (4); In the formula, represents the modulation index; represents the DC voltage; represents the sine function; represents the cosine function.

[0019] Among them, the specific process of obtaining the harmonic amplitude is as follows: In the process of calculating the harmonic amplitude of the SVPWM circuit, the point voltage is Fourier decomposed based on the combined output voltage vector, expressed as: (5); In the formula, represents the voltage waveform of the point voltage changing with time ; represents the average value of the point voltage; represents infinity; represents the harmonic order; represents the fundamental angular frequency; and respectively represent the amplitudes of the corresponding cosine term and sine term; among them, ; is the fundamental frequency; The total time of the positive and negative levels of the point voltage is the same within one output switching period T s , and the point voltage is an even function, expressed as: (6); (7); (8); In the formula, represents a small time increment; Taking the voltage space vector rotation angle as the boundary, the voltage waveform of the point voltage changing with time is expanded within half of the output switching period T s , divided into region 1 and region, and the lengths of region 1 and region are T s / 2; is the modulation factor; Based on the division into region 1 and region, the voltage space vector rotation angle corresponding to the kth region is set, expressed as: (9); In the formula, r k is the voltage space vector rotation angle corresponding to the kth region; represents the rotation angle increment of each interval; represents the total number of intervals; Let the positive level duty ratio of the voltage space vector rotation angle corresponding to the kth region be , expressed as: (10); wherein, represents the starting moment of the time node in half of the switching period T; s the starting moment of the time node; represents the last time node within half of the switching period T, that is, the end moment of half of the period T; s the end moment of half of the period T; s the end moment; represents the end moment of the positive level duty cycle time node corresponding to the voltage space vector rotation angle in the first zone; represents the positive level duty cycle of the voltage space vector rotation angle corresponding to the first zone; represents the end moment of the positive level duty cycle time node corresponding to the voltage space vector rotation angle in the (2k - 2)th zone; represents the end moment of the positive level duty cycle time node corresponding to the voltage space vector rotation angle in the (2k - 3)th zone; represents the positive level duty cycle of the voltage space vector rotation angle corresponding to the (k - 1)th zone; represents the end moment of the positive level duty cycle time node corresponding to the voltage space vector rotation angle in the (2k - 1)th zone; represents the period; is calculated based on the positive level duty cycle being and the period T to obtain the time value; represents the time value obtained from the positive level duty cycle in the first zone and the period T, represents the zone positive level duty cycle and the time value obtained from the period T; Then, the obtained time value is substituted into formula (8), and the integral of the time value is expanded to obtain , which represents: (11); wherein, represents the th harmonic amplitude.

[0020] Among them, the specific process of introducing the Levy tangent flight strategy into the black-winged kite optimization algorithm is as follows: Set the parameters of the Black-winged Kite Optimization Algorithm. The parameters include the number of black-winged kites, the search space, the total number of iterations, and the leader. The leader is a randomly selected value between [0, 1], which is used to initialize the position of each black-winged kite. Generate a black-winged kite matrix based on the set parameters of the Black-winged Kite Optimization Algorithm. After randomly assigning the position of each black-winged kite from the black-winged kite matrix, use the Tent mapping to generate the initial population, and set the current iteration number t of the initial population to 0. Calculate the initial position of the black-winged kites in the initial population through formula (12), which is expressed as: (12); In the formula, represents the initial position of the i-th black-winged kite; and represent the lower and upper bounds of the search space respectively; () represents a random number between [0, 1]; When initializing the initial population, regard the black-winged kite with the best fitness as the leader in the initial population, that is, the optimal position of the black-winged kites in the initial population. The optimal position is used as the best choice for the entire data, y L The calculation formula of is expressed as: (13); (14); In the formula, represents the fitness value of the currently found optimal solution; represents the fitness value of the i-th black-winged kite in the initial population; represents the optimal position of the black-winged kites in the initial population; represents the black-winged kite; represents function; represents the minimum value; represents the i-th black-winged kite at the current position at the fitness value; As a predator of small grassland mammals and insects, the black-winged kite adjusts the angles of its wings and tail according to the wind speed during combat, hovers quietly to observe prey, and then dives quickly to attack. The attack strategy includes global exploration and different attack behaviors, which help the black-winged kite better capture prey. The attack strategy update equation is: (15); In the formula, represents the position of the i-th black-winged kite in the j-th dimension at the (t + 1)-th iteration step; represents the position of the i-th black-winged kite in the j-th dimension at the t-th iteration step; represents the threshold; represents a random number between 0 and 1 in the initial population; represents a non - linear factor; ; Q represents the total number of iterations; represents the base of the natural logarithm; represents otherwise; The bird migration strategy is a complex behavior affected by environmental factors such as climate and food supply; the bird migration strategy is to adapt to seasonal changes. Many birds migrate from the north to the south in winter to obtain better living conditions and resources; the migration strategy is usually led by a leader, and the navigation skills of birds are crucial for the success of the group; based on the bird migration strategy, a hypothesis is proposed: if the fitness value of the currently found optimal solution is less than that of the randomly migrating population, the leader will abandon leadership and join the randomly migrating population, indicating that it is not suitable to lead the initial population forward. On the contrary, if the fitness value of the currently found optimal solution is greater than that of the randomly migrating population, it will guide the initial population until the destination is reached; the migration strategy can dynamically select excellent leaders to ensure a successful migration; the migration strategy of the black - winged kite is expressed as: (16); In the formula, represents a random number between 0 and 1, representing Cauchy mutation; represents the one with the highest score of the black - winged kite in the j - th dimension at the t - th iteration step; represents the fitness value of the i - th black - winged kite; represents a randomly generated fitness threshold; represents a constant; ; The probability density function of Cauchy mutation is expressed as: (17); In the formula, represents the probability density function of the Cauchy distribution; represents the scale parameter; represents a random variable; represents the location parameter; When the probability density function becomes the standard form, which is expressed as: (18); In the migration strategy stage, the Levy tangent flight strategy algorithm is proposed for the case where the fitness value of the currently found optimal solution is less than that of the randomly migrating population, which can get rid of the dependence on the black - winged kite that has lost its leadership role and is beneficial to expanding the search range of the algorithm; A new step size based on the tangent function is proposed in the Levy tangent flight strategy algorithm, such as: step xtan(0), where step is the moving step size and xtan(0) is the moving direction, which can finely adjust the local search; when the value of step xtan(0) gradually approaches 0, the tangent slope of the value of step xtan(0) gradually becomes smaller, the change of the moving step size is smaller, and the obtained data is more in line with the required optimized data, which is suitable for local search; when the value of step xtan(0) is closer to 0, the tangent slope of the value of step xtan(0) is larger, the change of the moving step size is larger, and the obtained solution is farther from the current solution, which is suitable for global search; using the Levy tangent flight strategy algorithm to generate the moving step size of tangent flight helps to explore the long-distance areas in the space, enhances the global search ability of the black-winged kite optimization algorithm, and can better cope with the complex search space at the same time. Adding the Levy tangent flight strategy algorithm to the black-winged kite migration strategy means: (19); In the formula, represents the tangent function; represents the flight length.

[0021] Among them, the specific process of introducing the prey escape strategy into the black-winged kite optimization algorithm is as follows: In the attack strategy of the black-winged kite optimization algorithm, it includes different attack behaviors of global exploration and search, but ignores the situation that the prey will escape during the attack of the black-winged kite, resulting in the failure of the black-winged kite optimization algorithm to find the global optimal solution and easily leading to the black-winged kite optimization algorithm falling into the local optimum. In addition, the search intensity of the black-winged kite optimization algorithm is limited due to the lack of information exchange during the attack stage. Therefore, referring to the prey escape energy in the golden jackal optimization algorithm, the prey escape energy in the black-winged kite algorithm is defined as follows: (20); In the formula, represents the prey escape energy; represents the exponential function;<s represents the ratio of the current iteration number t to the total iteration number ; When[[ID=?]] it means that the prey does not have enough prey escape energy to escape the attack circle of the black-winged kite. At this time, the original attack strategy of formula (15) is used to update the position of the black-winged kite, where E0 represents the prey escape threshold; Compared with formula (15), the non-linear factor becomes E, and the non-linear factor gradually decreases with the increase of the iteration number, and the non-linear factor It should be noted that there seems to be an unclear tag " " in the original text which might need further clarification for a more accurate translation.The limited range of value changes results in a limited search range for the Black-winged Kite Optimization Algorithm. And the oscillation of E between (-1, 1) makes the update step size of the Black-winged Kite Optimization Algorithm more random, enabling the algorithm to search for and find more excellent solutions; When it means that the prey has enough escape energy to escape the attack of the black-winged kite. At this time, a trust region mutation strategy is introduced in the Black-winged Kite Optimization Algorithm to increase the search intensity and prevent the prey from escaping; First, define the Euclidean distance between black-winged kite i and black-winged kite Based on the Euclidean distance, calculate the mean position of all black-winged kites , and calculate the initial trust region radius according to the distance between the current black-winged kite i and the mean position , which is expressed as: (21); In the formula, represents the Euclidean distance between black-winged kite i and black-winged kite ; represents the Euclidean distance; represents the coordinate value of the i-th black-winged kite on the first dimension in the multi-dimensional space; represents the -th black-winged kite's coordinate value on the first dimension in the multi-dimensional space; represents the coordinate value of the i-th black-winged kite on the D-th dimension in the multi-dimensional space; represents the -th black-winged kite's coordinate value on the D-th dimension in the multi-dimensional space; represents the total number of black-winged kites; represents the positions of black-winged kite i and black-winged kite ; represents the initial trust region radius of the i-th black-winged kite; If the Euclidean distance between the current black-winged kite i and black-winged kite is less than the initial trust region radius of the i-th black-winged kite, it is regarded as a trustworthy black-winged kite, and calculate the mean position of the trustworthy black-winged kites, which is expressed as: (22); In the formula, represents the set of positions of the i-th trustworthy black-winged kite; represents the position of the i-th trustworthy black-winged kite; represents the mean position of the trustworthy black-winged kites; represents the number of trustworthy black-winged kites; represents the threshold of the mean position of the i-th trustworthy black-winged kite; Finally, combining the information between the current trustworthy black-winged kites and the mean position of the trustworthy black-winged kites, the next generation of trustworthy black-winged kites is jointly generated to complete the information exchange between the trustworthy black-winged kites and prevent the black-winged kite optimization algorithm from falling into a local optimum; Introduce the Levy tangent flight strategy into the attack strategy. The Levy tangent flight strategy depicts a random search method, enabling the black-winged kite optimization algorithm to better perform local exploration at small step sizes and global [0,1] search at large step sizes; Introduce the Levy tangent flight strategy into the attack strategy, which means: (23); (24); In the formula, represents the step size of the Levy tangent flight strategy algorithm; represents a random number that follows a normal distribution between [0,1]; represents the standard deviation; represents the gamma function; represents a conventional parameter; In summary, combining the Levy tangent flight strategy and the trust region mutation strategy to update the position of the attack strategy, which means: (25); The error of the position of the j-th dimension and the i-th black-winged kite in the (t + 1)-th iteration after combination is corrected according to the optimal position y L of the black-winged kites in the initial population until the accuracy requirement of the black-winged kite optimization algorithm is met; When the black-winged kite optimization algorithm reaches the accuracy requirement, the iteration ends and the optimal identification parameters are obtained. Otherwise, return to formula (12) to recalculate the initial position of the black-winged kites in the initial population. Considering that when the number of iterations of the black-winged kite optimization algorithm reaches the upper limit during the iteration process and still does not meet the accuracy requirement, the minimum fitness value of the position of the j-th dimension and the i-th black-winged kite in the (t + 1)-th iteration after combination during the iteration process is selected as the optimal solution for output.

[0022] Among them, the specific process of obtaining the total harmonic distortion rate is as follows: Taking the solution of 6 switching tubes when M = 0.6 as an example, compare the solution results of the particle swarm optimization algorithm PSO, the black-winged kite optimization algorithm BKA, and the improved dynamic search algorithm EBKA; Iteratively calculate the th harmonic amplitude through the optimal identification parameters of the improved dynamic search algorithm, which means: (26); In the formula, represents the first harmonic amplitude; Indicates the distortion rate of the first 50 harmonics of the output voltage; Indicates the error term of the harmonic amplitude; Indicates the pulse edge; Indicates the i-th switching device; Is the amplitude of the zero-order harmonic; Indicates the error term of the amplitude of the first harmonic; Z indicates the modulation ratio; Indicates voltage; Indicates the correction factor or scale factor; Indicates the standard of the 5th harmonic amplitude; Indicates the error term of the 5th harmonic amplitude; Indicates the th harmonic amplitude standard; Indicates the th harmonic amplitude error term; = 1, 2, 3, …, 6; = 5, 7, 11, …, 49; Based on the pulse edge The pulse edge The value at the rising edge is 1, and the value at the falling edge of the pulse edge is -1 and the harmonic standards of each order of the power quality public power grid harmonics (GB / T 14549-93); Based on the pulse edge and the harmonic standards of each order of the power quality public power grid harmonics (GB / T 14549-93), the particle swarm optimization algorithm PSO, the black-winged kite optimization algorithm BKA and the improved dynamic search algorithm EBKA are used to repeatedly test the switching devices, and the optimal situation of the repeated test solution results is selected and applied to the specific harmonic elimination pulse width modulation and the specific harmonic suppression pulse width modulation strategies respectively; Based on the distortion rate of the first 50 harmonics of the output voltage Calculate the total harmonic distortion rate: (27); In the formula, Indicates the total harmonic distortion rate; Indicates the absolute value of the harmonic amplitude; Under the specific harmonic elimination pulse width modulation strategy, with the nonlinear factor With the increase in quantity, the convergence performance of the improved dynamic search algorithm EBKA is significantly better than that of the particle swarm optimization algorithm PSO and the black-winged kite optimization algorithm BKA. The search speed of the improved dynamic search algorithm EBKA increases at the initial stage, slows down at the mature stage, and speeds up at the final stage, similar to the growth trend of a growth curve, which is more conducive to jumping out of the local optimal solution and searching for the global optimal solution in the shortest time. The total harmonic distortion rate THD calculated for the selected switching tubes is also the smallest, while the particle swarm optimization algorithm PSO and the black-winged kite optimization algorithm BKA show oscillatory instability in the later stage of the search; Under the same modulation ratio M = 0.6, the improved dynamic search algorithm EBKA, the black-winged kite optimization algorithm BKA, and the particle swarm optimization algorithm PSO are respectively applied to the specific harmonic elimination pulse width modulation strategy and the specific harmonic suppression pulse width modulation strategy, and the obtained switching tubes and total harmonic distortion rates are shown in Tables 1 and 2.

[0023] Table 1 Comparison of simulation results of specific harmonic elimination pulse width modulation strategy under three algorithms Through the verification in Table 1, it is found that the optimal switching tubes obtained by the improved dynamic search algorithm EBKA are used to control the rectifier circuit. The total harmonic distortion rate THD of the rectifier circuit output is reduced by 1.97% compared with the particle swarm optimization algorithm PSO and by 0.73% compared with the black-winged kite optimization algorithm BKA.

[0024] Table 2 Comparison of simulation results of specific harmonic suppression pulse width modulation strategy under three algorithms Based on the improved dynamic search algorithm EBKA, taking the specific harmonic elimination pulse width modulation strategy and the specific harmonic suppression pulse width modulation strategy as examples, a fast Fourier transform FFT analysis is performed on the total harmonic distortion rate THD, which improves the overall iterative search ability and reduces the total harmonic distortion rate THD of the output voltage.

[0025] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for optimizing the active suppression of harmonics in a rectifier transformer for hydrogen production by water electrolysis, characterized in that: It includes the following steps: Step S1: Construct a rectifier circuit. The rectifier circuit includes a switching tube. Combine the voltage vectors and zero vectors output by the switching tube to obtain a combined output voltage vector. Process the combined output voltage vector to obtain the harmonic amplitude; Step S2: Introduce the Levy tangent flight strategy and the trust region mutation strategy into the black-winged kite optimization algorithm to construct an improved dynamic search algorithm; Step S3: Iteratively calculate the harmonic amplitude through the improved dynamic search algorithm to obtain the total harmonic distortion rate, that is, the optimal switching tube combination. Adjust the switching tubes of the rectifier circuit through the optimal switching tube combination to reduce the total harmonic distortion rate of the rectifier circuit.

2. An active harmonic suppression optimization method for a rectifier transformer in water electrolysis hydrogen production according to claim 1, characterized in that: The rectifier circuit includes an SVPWM circuit; the SVPWM circuit is divided into several sectors, each sector contains switching tubes, and also includes a voltage space vector rotation angle. The voltage vectors and zero vectors output by the switching tubes in several sectors are combined to obtain a combined output voltage vector; r is the voltage space vector rotation angle; is the combined output voltage vector; among them, the SVPWM circuit includes the midpoint voltage of the a-phase capacitor; is the point voltage.

3. An active harmonic suppression optimization method for a rectifier transformer in water electrolysis hydrogen production according to claim 2, characterized in that: Based on the combined output voltage vector, let the switching period be T s , which means: （1）； Wherein, ; represents a zero vector; represents a first zero vector; represents a second zero vector; represents the acting time of the first zero vector U1; represents the second zero vector acting time; represents the acting time of the zero vector acting time; Solve for the values of T1 and T2, expressed as: （2）； （3）； （4）； In the formula, represents the modulation index; represents the DC voltage; represents the sine function; represents the cosine function.

4. An active harmonic suppression optimization method for a rectifier transformer in water electrolysis hydrogen production according to claim 3, characterized in that: The specific process of obtaining the harmonic amplitude is as follows: In the process of calculating the harmonic amplitude of the SVPWM circuit, Fourier decomposition is performed on the point voltage based on the combined output voltage vector, expressed as: （5）； Wherein, represents the voltage waveform of the point voltage varying with time ; represents the average value of the point voltage; represents infinity; represents the harmonic order; represents the fundamental angular frequency; and respectively represent the amplitudes corresponding to the cosine term and the sine term; wherein ; is the fundamental frequency; The point voltage is an even function, expressed as: （6）； （7）； （8）； In the formula, represents a tiny time increment; Taking the voltage space vector rotation angle as the boundary, the voltage waveform of the point voltage changing with time is expanded within half of the output switching period T and divided into Region 1 and s Region; Region; is the modulation factor; Based on being divided into Zone 1 and the rotation angle of the voltage space vector corresponding to the k-th zone in the zone is represented as: （9）； where r k is the rotation angle of the voltage space vector corresponding to the k-th zone; represents the rotation angle increment of each interval; represents the total number of intervals; Let the positive-level duty ratio of the voltage space vector rotation angle corresponding to the k-th region be , which means: （10）； Wherein, represents half of the switching period T s starting time of the time node; represents half of the switching period T s the last time node within, that is, the end time of half of the period T s the end time; represents the end time of the positive level duty ratio time node of the voltage space vector rotation angle corresponding to the first region; represents the positive level duty ratio of the voltage space vector rotation angle corresponding to the first region; represents the end time of the positive level duty ratio time node of the voltage space vector rotation angle corresponding to the 2k - 2 region; represents the end time of the positive level duty ratio time node of the voltage space vector rotation angle corresponding to the 2k - 3 region; represents the positive level duty ratio of the voltage space vector rotation angle corresponding to the k - 1 region; represents the end time of the positive level duty ratio time node of the voltage space vector rotation angle corresponding to the 2k - 1 region; represents the period; Based on the positive level duty cycle And the period T is calculated to get Time value; Indicates the positive level duty cycle of zone 1 The time value obtained by and period T, Indicates the Positive level duty cycle The time value obtained by summing the period T; Then substitute the obtained time value into formula (8), and expand the integral of the time value to obtain , which represents: （11）； In the formula, represents the amplitude of the nth harmonic.

5. An active harmonic suppression optimization method for a rectifier transformer in water electrolysis hydrogen production according to claim 4, characterized in that: The specific process of introducing the Levy tangent flight strategy into the black-winged kite optimization algorithm is as follows: Set the parameters of the black-winged kite optimization algorithm. The parameters include the number of black-winged kites, the search space, the total number of iterations, and the leader. Generate a black-winged kite matrix based on the set parameters of the black-winged kite optimization algorithm. After randomly assigning the positions of each black-winged kite from the black-winged kite matrix, use Tent mapping to generate the initial population, and use formula (12) to calculate the initial positions of the black-winged kites in the initial population, expressed as: （12）； Wherein, represents the initial position of the i-th black-winged kite; and respectively represent the lower bound and the upper bound of the search space; () represents a random number between [0, 1]; The optimal position y of the black-winged kite in the initial population L The calculation formula of, indicating: （13）； （14）； In the formula, represents the currently found optimal fitness value; represents the fitness value of the i-th black-winged kite in the initial population; represents the black-winged kite; represents function; represents the minimum value; represents the fitness value of the i-th black-winged kite at the current position here; The attack strategy update equation is: （15）； In the formula, represents the position of the \(i\)-th black-winged kite in the \(j\)-th dimension at the \((t + 1)\)-th iteration step; represents the position of the \(i\)-th black-winged kite in the \(j\)-th dimension at the \(t\)-th iteration step; represents the threshold; represents a random number between 0 and 1 in the initial population; represents the non-linear factor; ; \(Q\) represents the total number of iterations; represents the base of the natural logarithm; represents otherwise; The black-winged kite migration strategy, expressed as: （16）； In the formula, represents a random number between 0 and 1, indicating the Cauchy mutation; represents the one with the highest score of the black-winged kite in the j-th dimension in the t-th iteration step; represents the fitness value of the i-th black-winged kite; represents a randomly generated fitness threshold; represents a constant; ; The probability density function of Cauchy mutation, expressed as: （17）； In the formula, represents the probability density function of the Cauchy distribution; represents the scale parameter; represents the random variable; represents the location parameter; When the probability density function becomes the standard form, indicating: （18）； Add the Levy tangent flight strategy algorithm to the black-winged kite migration strategy. The improved black-winged kite migration strategy, expressed as: （19）； In the formula, represents the tangent function; represents the flight length.

6. An active harmonic suppression optimization method for a rectifier transformer in water electrolysis hydrogen production according to claim 5, characterized in that: The specific process of introducing the trust region mutation strategy into the black-winged kite optimization algorithm is as follows: First, define the prey escape energy of the black-winged kite algorithm, expressed as: （20）； In the formula, represents the prey escape energy; represents the exponential function; represents the ratio of the current iteration number t to the total iteration number ; When it means that the prey does not have enough extra prey escape energy to escape from the attack circle of the black-winged kite. At this time, the original attack strategy of formula (15) is used to update the position of the black-winged kite, where E0 represents the prey escape threshold; When it means that the prey has enough prey escape energy to escape the attack of the black-winged kite. At this time, a trust region mutation strategy is introduced in the black-winged kite optimization algorithm to increase the search intensity and prevent the prey from escaping; First, define the Euclidean distance between black-winged kites i and black-winged kites and calculate the mean position of all black-winged kites based on the Euclidean distance , and calculate the initial trust region radius according to the distance between the current black-winged kite i and the mean position , which is expressed as: （21）； In the formula, represents the Euclidean distance between the black-winged kite i and the black-winged kite ; represents the Euclidean distance; represents the coordinate value of the i-th black-winged kite on the first dimension in the multi-dimensional space; represents the -th black-winged kite's coordinate value on the first dimension in the multi-dimensional space; represents the coordinate value of the i-th black-winged kite on the D-th dimension in the multi-dimensional space; represents the -th black-winged kite's coordinate value on the D-th dimension in the multi-dimensional space; represents the total number of black-winged kites; represents the positions of the black-winged kite i and the black-winged kite ; represents the initial trust region radius of the i-th black-winged kite; Current black-winged kite i to black-winged kite The Euclidean distance between them is less than the initial trust region radius of the i-th black-winged kite. As a trustworthy black-winged kite, calculate the mean position of the trustworthy black-winged kites, which is expressed as: （22）； In the formula, represents the set of positions of the i-th reliable black-winged kite; represents the position of the i-th reliable black-winged kite; represents the mean position of the reliable black-winged kites; represents the number of reliable black-winged kites; represents the threshold of the mean position of the i-th reliable black-winged kite; Introduce the Levy tangent flight strategy into the attack strategy, expressed as: （23）； （24）； In the formula, represents the step size of the Levy tangent flight strategy algorithm; represents a random number that follows a normal distribution between [0, 1]; represents the standard deviation; represents the gamma function; represents a conventional parameter; Combine the Levy tangent flight strategy and the trust region mutation strategy to update the position of the attack strategy, expressed as: （25）； The error of the position of the $i$-th black-winged kite in the $j$-th dimension at the $(t + 1)$-th iteration after combination is corrected according to the optimal position $y$ of the black-winged kites in the initial population L until the accuracy requirement of the black-winged kite optimization algorithm is met; When the black-winged kite optimization algorithm meets the accuracy requirements, end the iteration to obtain the optimal identification parameters. Otherwise, return to formula (12) to recalculate the initial positions of the black-winged kites in the initial population. Considering that when the number of iterations of the black-winged kite optimization algorithm reaches the upper limit during the iteration process and still does not meet the accuracy requirements, select the minimum fitness value of the position of the i-th black-winged kite in the j-th dimension at the (t + 1)-th iteration combined during the iteration process as the optimal solution output.

7. An active harmonic suppression optimization method for a rectifier transformer in water electrolysis hydrogen production according to claim 6, characterized in that: The specific process of obtaining the total harmonic distortion rate is as follows: Iterative calculation of the optimal identification parameter for the nth harmonic amplitude through an improved dynamic search algorithm , represented by; （26）； Wherein, represents the first harmonic amplitude; represents the total harmonic distortion rate of the first 50 harmonics of the output voltage; represents the error term of the harmonic amplitude; represents the pulse edge; represents the i-th switching device; is the zero harmonic amplitude; represents the error term of the first harmonic amplitude; Z represents the modulation ratio; represents the voltage; represents the correction factor or scale factor; represents the standard of the 5th harmonic amplitude; represents the error term of the 5th harmonic amplitude; represents the th harmonic amplitude standard; represents the th error term of the harmonic amplitude; = 1, 2, 3, …, 6; = 5, 7, 11, …, 49; Based on the first 50th harmonic distortion rate of the output voltage Calculate the total harmonic distortion rate, expressed as: （27）； Wherein, represents the total harmonic distortion rate; represents the absolute value of the harmonic amplitude.

Images