An improved meerkat algorithm optimization sapf current tracking control method

Abstract

The application discloses an improved honey badger algorithm optimization SAPF current tracking control method and relates to the technical field of power quality harmonic treatment. Traditional PR control can obtain large gain at the fundamental frequency, but the bandwidth is small. When the power grid frequency changes, the corresponding harmonic frequencies of each order will also change obviously and greatly, the current tracking control effect is affected, and the SAPF compensation result is poor. In order to improve the controller performance when the frequency changes, the SAPF current tracking control link in the application adopts a QPR control strategy. On the basis of optimizing the QPR control parameters by using the traditional particle swarm optimization (PSO), a new type of swarm intelligence algorithm, namely, the improved honey badger algorithm (GHBA), is introduced to set the control parameters of the QPR controller. Compared with the optimization result of the PSO algorithm, the harmonic distortion rate of the power grid current after the compensation of the control parameters of the QPR controller optimized by the GHBA algorithm is low, and the compensation requirement is met.

Description

Technical Field

[0001] This invention relates to the field of power quality harmonic mitigation technology, specifically to an improved honey badger algorithm-optimized SAPF current tracking control method. Background Technology

[0002] In recent years, with the acceleration of my country's industrialization, nonlinear and impulsive loads have been widely used, leading to increasingly serious harmonic current pollution of the power grid. SAPF (Short-terminal Filter) possesses high-performance harmonic suppression capabilities, and its current tracking control is one of the key technologies determining its compensation performance. SAPF current control requires controlling each harmonic current, which is more challenging in terms of accuracy and real-time performance compared to simply tracking the fundamental current. According to control theory, to track and control an input signal at a fixed frequency, the open-loop transfer function of the designed control system should have a very large gain at that frequency; this high gain is usually achieved by the designed controller. The current tracking control has a significant impact on the effectiveness of SAPF in compensating for harmonic currents, and the appropriate selection of control parameters is crucial for control performance.

[0003] For SAPF (Safety Assisted Power Source) systems, the reference current signal contains a large number of harmonic current components, mainly the 3rd, 5th, and 7th harmonics. To ensure efficient compensation and output a reasonable current compensation signal, PR (Positive Resonance) control is widely used in current tracking control strategies for SAPF systems. Traditional PR control can achieve a large gain at the fundamental frequency, but its bandwidth is small. When the grid frequency changes, even a small change, the corresponding harmonic frequencies will fluctuate significantly. This fails to amplify the gain for the specified harmonic error, affecting the current tracking control effect and leading to poor SAPF compensation results. To improve controller performance under frequency changes, the SAPF current tracking control stage adopts a QPR (Quick Resonance) control strategy.

[0004] Compared to PR controllers, QPR controllers have a larger amplitude gain at the resonant frequency and better tracking capability for the reference current command signal. Currently, the main parameter tuning methods include traditional parameter tuning methods and intelligent optimization algorithms. Traditional parameter tuning methods include the ZN method, the critical proportional gain method, and the attenuation curve method. These methods approximate tuning based on engineering empirical formulas and require final adjustments and refinements during actual operation. Their tuning accuracy is not high, and they require accurate object models, which are difficult to establish in industrial control. Intelligent algorithm optimization methods mainly include neural networks, fuzzy control, genetic algorithms, and particle swarm optimization algorithms. The tuning effect of neural networks is greatly affected by initial values; fuzzy control requires the tuner to have rich prior knowledge to write fuzzy rules; the crossover and mutation operations in genetic algorithms may worsen the optimal solution; particle swarm optimization is the most classic method among swarm intelligence algorithms, but it suffers from the drawbacks of easily getting trapped in local optima and slow convergence. Summary of the Invention

[0005] To address the aforementioned issues, this invention proposes using the improved Honey Badger Algorithm (GHBA) to tune the parameters of the SAPF current tracking QPR controller, and compares the results with those obtained using the traditional Particle Swarm Optimization (PSO) algorithm. Compared with the PSO algorithm, the grid current harmonic distortion rate is lower after the QPR controller's control parameters are compensated using the GHBA algorithm, and the SPAF exhibits better harmonic current compensation performance.

[0006] The present invention adopts the following technical solution:

[0007] Step 1, nonlinear load current i L Based on i p -i q The reference current detection module of the method, excluding the fundamental current i f Then, the harmonic current i is obtained. h .

[0008] Step 2, set the SAPF DC side voltage reference value U ref With the actual voltage value U dc The difference is fed into the PI controller, and the voltage deviation is converted into a compensation current increment Δi to stabilize the DC side voltage. p .

[0009] Step 3, adjust the compensation current increment Δi p With harmonic current i h By subtracting the reference current i, we can obtain the reference current i. ref .

[0010] Step 4, calculate the reference current i ref Subtract the compensation current i cThe difference is calculated, its absolute value is multiplied by time, and finally integrated. The result is used as the fitness function of the algorithm.

[0011] Step 5: Optimize the QPR control parameters using the improved honey badger algorithm GHBA to find K. p K r The optimal value is assigned to the QPR controller.

[0012] Step 6: Output signal g through current tracking control, i.e., QPR control.

[0013] Step 7: The output signal g controls the PWM converter so that the SAPF outputs a reasonable compensation current i. c This allows the compensation current to cancel out the harmonic current to be compensated in the load current, thereby reducing the grid current i s Approaching a sine wave.

[0014] Furthermore, the fitness function ITAE optimized in step 4 is calculated using the following method:

[0015] Current reference value i ref With output current value i c The difference is defined as e(t), then

[0016] Furthermore, in step 5, the improved honey badger algorithm GHBA includes the following steps:

[0017] Step 5.1: Initialize the honey badger population, setting the total number of honey badgers to N and the maximum number of iterations to t. max The upper and lower bounds of the search space are ub and lb, respectively. Within the defined boundary range, the positions of the honey badger population are randomly initialized as follows:

[0018] x ij =lb j +mod(x ij ,1)*(ub j -lb j )

[0019]

[0020] Where p is satisfied The smallest prime number, j = (1,2,…,D), where D is the dimension of the search space.

[0021] Step 5.2: Calculate the fitness value of all honey badgers in the population and determine the current optimal position X. prey .

[0022] Step 5.3, during each iteration, the density factor α is updated as follows:

[0023]

[0024] Where t represents the current iteration number, t max This represents the maximum number of iterations, where C is a constant of 2.

[0025] Step 5.4, during each iteration, the bee attraction I i The update description is as follows:

[0026]

[0027] S=(x i -x i+1 ) 2

[0028] d i =x prey -x i

[0029] Where S is the source intensity, d i x is the distance between the prey and the i-th honey badger, r2 is a random constant in [0,1], and x prey It is the current optimal position, x i It is the position of the i-th honey badger.

[0030] Step 5.5, Honey Badger position update: Based on the value of the control direction parameter F, it is determined whether the position update uses the digging stage or the honey stage. The update formula is as follows:

[0031]

[0032] x new =x prey +F*β*I*x prey +F*r3*α*d i *|cos(2πr⁴)*[1-cos(2πr⁵)]|

[0033] x new =x prey +F*r7*α*d i

[0034] Where r3, r4, r5, r6, and r7 are random constants in the range [0,1]. When r6 ≤ 0.5, the honey badger will choose the digging stage; otherwise, it will choose the honey stage. β represents the honey badger's ability to obtain food, and here it is taken as a constant value of 6.

[0035] Step 5.6: Determine if the termination condition is met. If it is, output the optimal solution and end the program; otherwise, repeat the above improved honey badger algorithm process to continue the optimization iteration.

[0036] Step 5.7, optimize the parameter K obtained by the improved honey badger algorithm. p K r Assign the value to the QPR controller.

[0037] Furthermore, in step 6, QPR control not only maintains the high gain of the PR controller but also has a larger bandwidth, effectively reducing the impact of grid frequency offset. The transfer function of QPR control is as follows:

[0038]

[0039] ω in the transfer function c Used to adjust the gain and bandwidth of the QPR controller at the resonant point, ω c The smaller the value of K, the closer the peak gain of QPR at resonance is to PR. p It is the proportionality coefficient, K r This is the gain value of the QPR controller at the resonant point, where ω0 is the fundamental angular frequency, and ω c ω is the cutoff angular frequency, and s is the differential operator.

[0040] The beneficial effects of this invention are:

[0041] Compared to the PSO algorithm for optimizing current tracking QPR control parameters, the GHBA algorithm proposed in this invention for QPR parameter tuning results in a lower THD value for the grid current harmonic distortion after SAPF compensation. This significantly improves the effectiveness and real-time performance of harmonic compensation, reduces steady-state error, and is of great significance for effectively controlling harmonics and improving power quality. Attached Figure Description

[0042] Figure 1 The schematic diagram shows the SAPF current tracking QPR controller optimized by the GHBA algorithm of this invention.

[0043] Figure 2 Optimize the QPR control parameter iteration curve for the PSO algorithm;

[0044] Figure 3 FFT analysis diagram of the grid current for SAPF current tracking QPR control after PSO algorithm optimization;

[0045] Figure 4 The GHBA algorithm of this invention optimizes the iteration curve of the QPR control parameters;

[0046] Figure 5 The above is an FFT analysis diagram of the power grid current after the optimization of the GHBA algorithm of this invention to track the QPR control of the SAPF current. Detailed Implementation

[0047] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0048] like Figure 1 As shown, one embodiment of the present invention discloses an improved honey badger algorithm for optimizing SAPF current tracking QPR controller parameters, comprising the following steps:

[0049] Step 1, nonlinear load current i L Based on i p -i q The reference current detection module of the method, excluding the fundamental current i f Then, the harmonic current i is obtained. h .

[0050] Step 2, set the SAPF DC side voltage reference value U ref With the actual voltage value U dc The difference is fed into the PI controller, and the voltage deviation is converted into a compensation current increment Δi to stabilize the DC side voltage. p .

[0051] Step 3, adjust the compensation current increment Δi p With harmonic current i h By subtracting the reference current i, we can obtain the reference current i. ref .

[0052] Step 4, calculate the reference current i ref Subtract the compensation current i c The difference is calculated, its absolute value is multiplied by time, and finally integrated. The result is used as the fitness function of the algorithm.

[0053] Step 5: Optimize the QPR control parameters using the improved honey badger algorithm GHBA to find K. p K r The optimal value is assigned to the QPR controller.

[0054] Step 6: Output signal g through current tracking control, i.e., QPR control.

[0055] Step 7: The output signal g controls the PWM converter so that the SAPF outputs a reasonable compensation current i. c This allows the compensation current to cancel out the harmonic current to be compensated in the load current, thereby reducing the grid current i s Approaching a sine wave.

[0056] Furthermore, the fitness function ITAE optimized in step 4 is calculated using the following method:

[0057] Current reference value i ref With output current value i cThe difference is defined as e(t), then

[0058] Furthermore, in step 5, the improved honey badger algorithm GHBA includes the following steps:

[0059] Step 5.1: Initialize the honey badger population, setting the total number of honey badgers to N and the maximum number of iterations to t. max The upper and lower bounds of the search space are ub and lb, respectively. Within the defined boundary range, the positions of the honey badger population are randomly initialized as follows:

[0060] x ij =lb j +mod(x ij ,1)*(ub j -lb j )

[0061]

[0062] Where p is satisfied The smallest prime number, j = (1,2,…,D), where D is the dimension of the search space.

[0063] Step 5.2: Calculate the fitness value of all honey badgers in the population and determine the current optimal position X. prey .

[0064] Step 5.3, during each iteration, the density factor α is updated as follows:

[0065]

[0066] Where t represents the current iteration number, t max This represents the maximum number of iterations, where C is a constant of 2.

[0067] Step 5.4, during each iteration, the bee attraction I i The update description is as follows:

[0068]

[0069] S=(x i -x i+1 ) 2

[0070] d i =x prey -x i

[0071] Where r2 is a random constant in [0,1], x prey It is the current optimal position, x iIt is the position of the i-th honey badger, S is the source intensity, and d i It is the distance between the prey and the i-th honey badger.

[0072] Step 5.5, Honey Badger position update: Based on the value of the control direction parameter F, it is determined whether the position update uses the digging stage or the honey stage. The update formula is as follows:

[0073]

[0074] x new =x prey +F*β*I*x prey +F*r3*α*d i *|cos(2πr⁴)*[1-cos(2πr⁵)]|

[0075] x new =x prey +F*r7*α*d i

[0076] Where r4, r5, r6, and r7 are random constants in the range [0,1]. When r6 ≤ 0.5, the honey badger will choose the digging stage; otherwise, it will choose the honey stage. β represents the honey badger's ability to obtain food, and here it is taken as a constant value of 6.

[0077] Step 5.6: Determine if the termination condition is met. If it is, output the optimal solution and end the program; otherwise, repeat the above improved honey badger algorithm process to continue the optimization iteration.

[0078] Step 5.7, optimize the parameter K obtained by the improved honey badger algorithm. p K r Assign the value to the OPR controller.

[0079] Furthermore, in step 6, QPR control not only maintains the high gain of the PR controller but also has a larger bandwidth, effectively reducing the impact of grid frequency offset. The transfer function of QPR control is as follows:

[0080]

[0081] ω in the transfer function c Used to adjust the gain and bandwidth of the QPR controller at the resonant point, ω c The smaller the value of K, the closer the peak gain of QPR at resonance is to PR. p It is the proportionality coefficient, K r This is the gain value of the QPR controller at the resonant point, where ω0 is the fundamental angular frequency, and ω c ω is the cutoff angular frequency, and s is the differential operator.

[0082] In this embodiment, the algorithm parameters are taken as follows:

[0083] PSO algorithm optimization QPR controller parameters: Particle swarm size SwarmSize = 20, dimension Dim = 2, maximum number of iterations MaxIter = 20.

[0084] GHBA algorithm optimizes QPR controller parameters: honey badger population size N=20, dimension Dim=2, maximum number of iterations t max =20, the upper and lower limits of the search space are ub=100 and lb=-100 respectively.

[0085] The improved Honey Badger Algorithm (GHBA) proposed in this invention is used to optimize SAPF current tracking control, and the optimization results are compared with those of the Particle Swarm Optimization (PSO) algorithm. Figures 2 to 5 The results show that, compared with the PSO algorithm for setting the current tracking QPR controller parameters, the GHBA algorithm proposed in this invention for QPR parameter setting results in a smaller THD value of the grid current harmonic distortion rate after SAPF compensation and a better harmonic compensation effect.

[0086] Finally, it should be noted that the above description is only for specific embodiments of the present invention. However, the present invention is not limited to the specific embodiments described above. Equivalent modifications and substitutions made to the present invention by those skilled in the art are also within the scope of the present invention. Therefore, all equivalent changes and modifications made without departing from the spirit and scope of the present invention are covered within the scope of the present invention.

Claims

1. An improved honey badger algorithm-optimized SAPF current tracking control method, characterized in that: Includes the following steps: Step 1, nonlinear load current i L Based on i p -i q The reference current detection module of the method removes the fundamental current. Then, the harmonic current i is obtained. h ; Step 2, set the SAPF DC side voltage reference value. Compared with actual voltage value The difference is fed into the PI controller, and the voltage deviation is converted into a compensation current increment to stabilize the DC side voltage. ; Step 3, adjust the compensation current increment. With harmonic current i h By subtracting the reference current i, we can obtain the reference current i. ref ; Step 4, calculate the reference current i ref Subtract the compensation current i c The difference is calculated, its absolute value is multiplied by time, and finally integrated. The result is used as the fitness function of the algorithm. Step 5: Optimize the QPR control parameters using the improved honey badger algorithm GHBA to find K. p K r The optimal value is assigned to the QPR controller; the improved honey badger algorithm GHBA includes the following steps: Step 5.1: Initialize the honey badger population, setting the total number of honey badgers to N, and the maximum number of iterations to [value missing]. The upper and lower limits of the search space are respectively , Within the defined boundary range, the locations of the honey badger population are randomly initialized as follows: ， ， Where p is satisfied The smallest prime number, j=1,2, D, where D is the dimension of the search space; Step 5.2: Calculate the fitness value of all honey badgers in the population and determine the current optimal position. ; Step 5.3, during each iteration, the density factor The update description is as follows: ， Where t represents the current iteration number, This represents the maximum number of iterations, where C is a constant of 2; Step 5.4, during each iteration, the bee attractiveness... The update description is as follows: ， ， ， in, It is the source strength. Is it prey and the first The distance between honey badgers It is a random constant in [0,1]. This is the current optimal position. It is the first The location of the honey badger; Step 5.5, Honey Badger position update: Based on the value of the control direction parameter F, it is determined whether the position update uses the digging stage or the honey stage. The update formula is as follows: ， ， ， in, It is a random constant in [0,1], when Honey badgers will choose the digging stage when digging, and conversely, they will choose the honey-making stage when scavenging. This represents the honey badger's ability to obtain food, which is taken as a constant value of 6 here; Step 5.6: Determine if the termination condition is met. If it is, output the optimal solution and end the program; otherwise, repeat the above improved honey badger algorithm process to continue the optimization iteration. Step 5.7: Assign the parameters Kp and Kr obtained by the improved honey badger algorithm to the QPR controller; Step 6: Output signal g through current tracking control, i.e., QPR control; Step 7: The output signal g controls the PWM converter so that the SAPF outputs a reasonable compensation current i. c This allows the compensation current to cancel out the harmonic current to be compensated in the load current, thereby reducing the grid current i s Approaching a sine wave.

2. The improved honey badger algorithm-optimized SAPF current tracking control method according to claim 1, characterized in that: In step 4, the algorithm fitness function ITAE is calculated using the following method: Current reference value With compensation current The difference is defined as e(t), then .

3. The improved honey badger algorithm-optimized SAPF current tracking control method according to claim 1, characterized in that: In step 6, the transfer function of QPR control is as follows: ， In the transfer function Used to adjust the gain bandwidth of the QPR controller at the resonant point. The smaller the value, the closer the peak gain of QPR at resonance is to PR; It is a proportionality coefficient. It is the gain value of the QPR controller at the resonant point. The fundamental angular frequency, ω is the cutoff angular frequency, and s is the differential operator.

Images