Improved grey wolf MPPT (maximum power point tracking) control method and device of photovoltaic array suitable for charging of energy storage power station

Abstract

The invention relates to an improved grey wolf MPPT (maximum power point tracking) control method and device suitable for a photovoltaic array for charging an energy storage power station. The method comprises the following steps: constructing an improved grey wolf algorithm THW-GWO-Pamp; o, a traditional grey wolf algorithm GWO and maximum power tracking are combined, the GWO algorithm is improved, and the improved GWO algorithm is combined with a traditional Pamp; combining the GWO algorithm with the GWO algorithm to form an improved grey wolf algorithm THW-GWO-Pamp; o; setting simulation working conditions, and designing operation data of the photovoltaic array under different working conditions; a simulation model of a photovoltaic power generation system is built based on Matlab / Simulink, and THW-GWO-Pamp is used; the O algorithm controls the duty ratio of the circuit through the MPPT controller; the simulation model of the photovoltaic power generation system is operated, and THW-GWO-Pamp is analyzed; the difference between the actual voltage, current and power and the ideal voltage, current and power under different conditions is calculated through the O algorithm, and the efficiency of different algorithms under the same and different conditions is compared and analyzed; the simulation model respectively simulates global maximum power point tracking under a uniform illumination condition and a shading condition, and THW-GWO-Pamp is verified; and the O algorithm tracks the accuracy and speed of the maximum power point under the complex illumination condition.

Description

Technical Field

[0001] The present invention relates to the technical field of inverter control, and particularly to an improved grey wolf MPPT control method and device for a photovoltaic array applicable to charging of an energy storage power station. Background Art

[0002] With the rapid development of renewable energy (especially photovoltaic and wind energy), and the increasing demand for energy security, environmental protection and energy transformation, energy storage technology has become an important part of realizing an efficient, reliable and green power system. As a clean and renewable power generation technology, photovoltaic power generation has unique advantages in terms of resource utilization rate, ecological environment, etc., and has received extensive attention from scholars at home and abroad.

[0003] The output of a photovoltaic array has a non-linear characteristic under a certain fixed working condition, and there is a peak value, i.e., the maximum power point, in its output power. To improve the efficiency of a photovoltaic power generation system and avoid output power loss, the maximum power point tracking (MPPT) algorithm of the photovoltaic array is particularly important. The traditional MPPT algorithm changes the output characteristics of the photovoltaic array by changing the structure of the photovoltaic module, converts the multi-peak of the output power into a single peak, and thus controls the output of the photovoltaic array to be at the maximum power point.

[0004] However, in real life, due to the influence of external obstacles, the photovoltaic array is under partial shading conditions, and there are multiple peaks in the output power. The maximum power point tracking of the photovoltaic array is more complex. The traditional MPPT algorithm may not be able to jump out of the local optimum, and has great limitations in terms of dynamic response and stable accuracy, thus unable to complete accurate maximum power point tracking, resulting in the inability of the inverter to achieve maximum power point tracking under complex illumination conditions and affecting the efficiency of the energy storage system. Summary of the Invention

[0005] The present invention aims to overcome the above-mentioned drawbacks of the prior art and provides an improved grey wolf MPPT control method for a photovoltaic array applicable to charging of an energy storage power station.

[0006] The present invention adjusts the traditional grey wolf algorithm (GWO) introduced based on swarm intelligence in terms of its convergence factor, weighted distance, position update weight, and jumping out of the local optimum, and combines it with the perturbation and observation method (P&O), thereby adjusting the strategy of the maximum power point tracking algorithm, enabling the energy storage inverter to operate at the maximum power point to the greatest extent and improving the efficiency of the energy storage inverter; in addition, the present invention integrates the advantages of different MPPT algorithms and has more advantages in terms of dynamic response and stable accuracy compared with a single MPPT control method.

[0007] To achieve the above object, the first aspect of the present invention relates to an improved grey wolf MPPT control method for a photovoltaic array applicable to charging an energy storage power station. The duty ratio of the Boost circuit is adjusted by an MPPT controller using an improved grey wolf algorithm THW-GWO-P&O to keep the output power of the photovoltaic array near the maximum power point. Denote the input of the photovoltaic array as solar irradiance and temperature, and the output as voltage, current, power, and efficiency. The method includes the following steps:

[0008] S1. Construct the improved grey wolf algorithm THW-GWO-P&O; combine the traditional grey wolf algorithm GWO with maximum power tracking, improve the GWO algorithm, adjust the convergence factor, and then adjust the weighted distance and position update weight to form an improved grey wolf algorithm THW-GWO led by two wolves. Finally, combine it with the traditional P&O algorithm to form the improved grey wolf algorithm THW-GWO-P&O;

[0009] S2. Set the simulation conditions and design the operating data of the photovoltaic array under different conditions. Among them, the conditions are divided into uniform illumination and partial shading conditions, and the operating data are the different solar irradiance and temperature of each single photovoltaic cell of the photovoltaic array;

[0010] S3. Build a simulation model of the photovoltaic power generation system in Matlab / Simulink, including components such as a photovoltaic array, an MPPT controller, a Boost boost circuit, and a load. The photovoltaic array is composed of multiple photovoltaic cells, and the THW-GWO-P&O algorithm is used to control the circuit duty ratio through the MPPT controller;

[0011] S4. Take the different solar irradiance and temperature of each single photovoltaic cell of the photovoltaic array as data input, and voltage, current, power, and efficiency as data output. Run the simulation model of the photovoltaic power generation system, observe the gap between the actual voltage, current, and power and the ideal voltage, current, and power of the THW-GWO-P&O algorithm under different conditions, and conduct a comparative analysis of the efficiency of different algorithms under the same and different conditions; S5. The constructed simulation model simulates the global maximum power point tracking under uniform illumination conditions and shading conditions respectively. Through comparative analysis with the P&O algorithm, the GWO algorithm, and the THW-GWO algorithm, verify the accuracy and speed of the THW-GWO-P&O algorithm in tracking the maximum power point under complex illumination conditions.

[0012] Preferably, the step S1 specifically includes:

[0013] S101. The GWO algorithm obtains the optimal solution. The GWO algorithm is based on the principles of search, encirclement, and hunting. Grey wolves continuously iterate by adjusting their positions, gather towards the direction of the optimal solution, and finally achieve the goal;

[0014] S102. Formulate the optimization rules of the GWO algorithm. After calculating the fitness of the grey wolves each time, compare among the alpha wolf, beta wolf, and delta wolf to determine the new leader. The update process is as follows: First, compare the average fitness of the grey wolves with the alpha wolf. If it exceeds, it becomes the new alpha wolf; otherwise, it remains unchanged. Then, compare the fitness with the beta wolf. If it exceeds, replace the beta wolf. If the fitness is lower than that of the alpha wolf and beta wolf, compare it with the delta wolf. If it exceeds, it becomes the new delta wolf. If the fitness of the grey wolf is lower than that of the alpha wolf, beta wolf, and delta wolf, the leadership hierarchy remains unchanged.

[0015] S103. Obtain the global maximum power. Use the real-time power of the photovoltaic array as the fitness function, and the position of the grey wolf corresponds to the duty cycle. With each round of update, the grey wolf group gradually approaches the global maximum power point. By setting different input data of the photovoltaic array, the operating data of the GWO algorithm under different working conditions can be obtained. In this process, the fitness of wolf alpha is considered the maximum output power.

[0016] Preferably, the step S103 specifically includes:

[0017] S1031. Calculate the efficiency of the GWO algorithm. Under the current uniform and partial shading illumination conditions, record the differences between the actual voltage, current, power and the ideal voltage, current, power of the GWO algorithm under different conditions, and judge the level of efficiency.

[0018] S1032. Extract the change characteristics of the GWO algorithm. Study the unique global maximum power point among the multiple peaks of the tracking characteristic curve when the solar photovoltaic cell is under partial shading conditions, and extract the change characteristics between the ideal power and the actual power.

[0019] Preferably, the step S1 further includes:

[0020] S104. Improve the convergence factor and weighted distance. The global and local search capabilities of the GWO algorithm are both affected by the parameter A. When |A| ≤ 1, the wolf group tends to follow the alpha wolf for hunting. When |A| > 1, the wolf group tends to disperse to find prey. The parameter A changes with the size of the convergence factor a. Improve the GWO algorithm, and through the position update weight, fuse the problem that the balance between global exploration and local development may be poor when the GWO algorithm changes linearly.

[0021] S105. Introduce non-linear double convergence factors. In the GWO algorithm, the control algorithm for the first half of the iteration is used for global exploration of the optimal value, and the second half of the iteration is used for local search for the optimal value. Although the convergence factor a changes linearly, it may lead to a situation where the global and local optimization cannot reach a balance. To solve this problem, THW-GWO proposes a new method. First, calculate the average fitness value of the wolf pack. Wolves with a fitness value higher than the average are hunting wolves, while those lower than the average are scouting wolves. According to the classification, determine the corresponding non-linear double convergence factors. The improvement of the convergence factor ensures that the convergence factor a1 decreases slowly in the early and late stages and rapidly in the middle stage, achieving a balance between global search and local development, thereby improving the convergence speed of the algorithm. And the convergence factor a2 decreases non-linearly to 0, enhancing the global exploration ability of the scouting wolves;

[0022] S106. Introduce the global leading wolf γ to update the global optimal value and improve the GWO algorithm. By introducing the adaptive weight coefficient and Levy flight strategy, the problem that the GWO algorithm may lead to a poor balance between global exploration and local development during linear change is solved. Denote the global leading wolf γ as the global optimal value during the iteration process. The work of the γ wolf is to communicate with the hunting wolves, update the global optimal solution, and guide the scouting wolves to search for prey. Denote the position of the γ wolf as Pbest, and update the position. Levy flight is used in the update formula. And to improve the search efficiency of the scouting wolves, when they get a poor position after Levy flight, the position is not changed, and the position of the scouting wolves is updated again.

[0023] Preferably, the step S104 specifically includes:

[0024] Adopt a new jumping-out mechanism to avoid the dilemma caused by the wrong judgment of the leading hunting α wolf in the GWO algorithm. When the position of the α wolf remains unchanged after continuous iteration for n times, let the wolf pack perform a Levy flight to continue searching for other optimal solutions until the algorithm ends.

[0025] Preferably, the step S106 specifically includes:

[0026] Introduce the adaptive weight coefficient of the two-headed wolf mechanism. The GWO algorithm uses the average weight update coefficient, but this strategy will greatly slow down the hunting speed. The THW-GWO algorithm introduces the two-headed wolf mechanism, which is responsible for hunting wolves and scouting wolves respectively. Since the goal of hunting wolves is to surround and kill prey, while the task of scouting wolves is to search for prey. Therefore, the THW-GWO algorithm proposes a strategy that the moving weight of hunting wolves is determined by the speed of fitness value decrease. This strategy avoids the concentration of gray wolf individuals towards the α wolf, balances the influence between the α wolf, β wolf, and δ wolf, thereby enhancing the exploration ability of the algorithm.

[0027] Preferably, step S2 specifically includes:

[0028] Condition 1, set under uniform light conditions.

[0029] Condition 2, set under partial shading conditions.

[0030] Simulation information setting: During simulation, four aspects including photovoltaic output power, output current, output voltage, and photoelectric conversion efficiency are selected for analysis. The MPPT tracking efficiency is defined as the ratio of the maximum output power actually tracked by the photovoltaic system to the maximum output power theoretically achievable under the same environmental conditions (solar irradiance and temperature). A simulation model of the photovoltaic power generation system is built in Matlab / Simulink, and the photovoltaic array adopts a 5×1 series structure. Two simulation conditions are set during simulation, and the temperature is always maintained at 25°C. The control effects of the photovoltaic array under four MPPT methods, namely P&O, GWO, THW-GWO, and THW-GWO-P&O, are compared and analyzed, and the specific applications of each algorithm are realized through the S-function module.

[0031] Preferably, step S5 specifically includes:

[0032] S501. Initialize the algorithm, initialize the positions of 10 wolves, and these positions are evenly distributed between 0.1 and 1, representing the duty cycle of the photovoltaic array.

[0033] S502. Evaluate the fitness of the algorithm, select the output power P as the fitness function of the photovoltaic array, and evaluate the performance of each wolf. Select the three wolves with the top fitness rankings and use their position information to guide other grey wolves.

[0034] S503. Update the positions of the algorithm, update the positions of the wolves according to a specific formula, and make them approach the maximum power point. In this stage, the global search performance of the THW-GWO algorithm is utilized to quickly converge the wolf pack near the maximum power point.

[0035] S504. Conduct local search for the algorithm. When the maximum number of iterations is reached or close to the MPP, a P&O algorithm with a small step size is used for local search. The P&O algorithm has good local search performance and fast convergence performance, and can achieve precise adjustment near the maximum power point. In this stage, a P&O algorithm with a small step size is used for local search until the most accurate MPP position is determined. When the termination condition is met, the optimal value and the optimal duty cycle are output.

[0036] The second aspect of the present invention relates to an improved grey wolf MPPT control device for a photovoltaic array applicable to charging of an energy storage power station, including a memory and one or more processors. Executable code is stored in the memory. When the one or more processors execute the executable code, it is used to implement the improved grey wolf MPPT control method for a photovoltaic array applicable to charging of an energy storage power station according to the present invention.

[0037] The third aspect of the present invention relates to a computer-readable storage medium, on which a program is stored. When the program is executed by a processor, it implements the improved grey wolf MPPT control method for a photovoltaic array applicable to charging of an energy storage power station according to the present invention.

[0038] Compared with the prior art, the advantages of the present invention are as follows: taking the different solar irradiances and temperatures of each single photovoltaic cell in the photovoltaic array as data inputs, and taking voltage, current, power, and efficiency as data outputs, running the simulation model of the photovoltaic power generation system, observing the gap between the actual voltage, current, power and the ideal voltage, current, power of the THW-GWO-P&O algorithm under different conditions, and conducting a comparative analysis of the efficiencies of different algorithms under the same and different conditions. The constructed simulation model respectively simulates the global maximum power point tracking under uniform illumination conditions and shading conditions. Through comparative analysis with the P&O algorithm, GWO algorithm, and THW-GWO algorithm, the accuracy and speed of the THW-GWO-P&O algorithm for tracking the maximum power point under complex illumination conditions are verified. By introducing the traditional grey wolf algorithm GWO based on swarm intelligence, adjusting its convergence factor, weighted distance, position update weight, and jumping out of local optimum, etc., and combining it with the perturbation observation method P&O, the strategy of the maximum power point tracking algorithm is adjusted, so that the energy storage inverter operates at the maximum power point to the greatest extent, improving the efficiency of the energy storage inverter; in addition, the present invention integrates the advantages of different MPPT algorithms and has more advantages in terms of dynamic response and stable accuracy compared with a single MPPT control method. Description of the Drawings

[0039] Figure 1 is a flowchart of the method of the present invention;

[0040] Figure 2 is a topology diagram of the inverter and photovoltaic power generation system of the present invention;

[0041] Figure 3 is an output characteristic curve diagram of the photovoltaic array under partial shading of the present invention;

[0042] Figure 4 is a schematic diagram of the search process of the GWO algorithm of the present invention;

[0043] Figure 5is the flowchart of the THW - GWO - P&O composite MPPT control of the photovoltaic array of the present invention;

[0044] Figure 6 is the schematic diagram of the device of the present invention. Specific embodiments

[0045] To describe in detail the technical content, structural features, achieved objectives and effects of the present invention, the following is a detailed description in conjunction with the embodiments and with reference to the accompanying drawings.

[0046] Embodiment 1

[0047] Please refer to Figures 1 - 5 As shown, the improved grey wolf MPPT control method for a photovoltaic array applicable to charging an energy storage power station in this embodiment is characterized in that the duty ratio of the Boost circuit is adjusted by an MPPT controller using the improved grey wolf algorithm THW - GWO - P&O to keep the output power of the photovoltaic array near the maximum power point. Denote the input of the photovoltaic array as solar irradiance and temperature, and the output as voltage, current, power, and efficiency. The method includes the following steps:

[0048] S1. Construct the improved grey wolf algorithm THW - GWO - P&O, which combines the traditional grey wolf algorithm GWO with maximum power tracking, improves the GWO algorithm, adjusts the convergence factor, and then adjusts the weighted distance and position update weight to form the improved grey wolf algorithm THW - GWO led by two wolves. Finally, it is combined with the traditional P&O algorithm to form the improved grey wolf algorithm THW - GWO - P&O;

[0049] S2. Set the simulation conditions and design the operating data of the photovoltaic array under different conditions. Among them, the conditions are divided into uniform illumination and partial shading conditions, and the operating data are the different solar irradiances and temperatures of each single photovoltaic cell of the photovoltaic array;

[0050] S3. Build a simulation model of a photovoltaic power generation system in Matlab / Simulink, including components such as a photovoltaic array, an MPPT controller, a Boost boost circuit, and a load. The photovoltaic array is composed of multiple photovoltaic cells, and the THW - GWO - P&O algorithm is used to control the circuit duty ratio through the MPPT controller;

[0051] S4. Taking the different solar irradiances and temperatures of each individual photovoltaic cell in the photovoltaic array as data inputs, and voltage, current, power, and efficiency as data outputs, run the simulation model of the photovoltaic power generation system, observe the gap between the actual voltage, current, power and the ideal voltage, current, power under different conditions of the THW-GWO-P&O algorithm, and conduct a comparative analysis of the efficiencies of different algorithms under the same and different conditions; S5. The constructed simulation model simulates the global maximum power point tracking under uniform illumination conditions and shading conditions respectively. By comparing and analyzing with the P&O algorithm, GWO algorithm, and THW-GWO algorithm, verify the accuracy and speed of the THW-GWO-P&O algorithm in tracking the maximum power point under complex illumination conditions.

[0052] Preferably, the step S1 specifically includes:

[0053] S101. The GWO algorithm obtains the optimal solution. The GWO algorithm is based on the principles of search, encirclement, and hunting. The grey wolves continuously iterate by adjusting their positions and gather towards the direction of the optimal solution, ultimately achieving the goal.

[0054] S102. Formulate the optimization rules of the GWO algorithm. After calculating the fitness of the grey wolves each time, compare among the alpha wolf, beta wolf, and delta wolf to determine the new leader. The update process is as follows: First, compare the average fitness of the grey wolves with the alpha wolf first. If it exceeds, it becomes the new alpha wolf, otherwise it remains unchanged. Then, compare the fitness with the beta wolf. If it exceeds, replace it. If the fitness is less than that of the alpha wolf and beta wolf, compare it with the delta wolf. If it exceeds, it becomes the new delta wolf. If the fitness of the grey wolf is lower than that of the alpha wolf, beta wolf, and delta wolf, the leadership hierarchy remains unchanged.

[0055] S103. Obtain the global maximum power. Take the real-time power of the photovoltaic array as the fitness function, and the position of the grey wolf corresponds to the duty ratio. With each round of update, the grey wolf group gradually approaches the global maximum power point. By setting different input data for the photovoltaic array, obtain the operation data of the GWO algorithm under different working conditions. In this process, the fitness of the alpha wolf is considered to be the maximum output power.

[0056] Preferably, the step S103 specifically includes:

[0057] S1031. Calculate the efficiency of the GWO algorithm. Under the current uniform and partial shading illumination conditions, record the gap between the actual voltage, current, power and the ideal voltage, current, power of the GWO algorithm under different conditions, and judge the level of efficiency.

[0058] S1032. Extract the change characteristics of the GWO algorithm. Study the unique global maximum power point among the multiple peaks of the tracking characteristic curve of the GWO algorithm when the solar photovoltaic cell is in the condition of partial shading, and extract the change characteristics between the ideal power and the actual power.

[0059] Preferably, the step S1 further includes:

[0060] S104. Improve the convergence factor and weighted distance. The global and local search capabilities of the GWO algorithm are both affected by the parameter A. When |A| ≤ 1, the wolf pack tends to follow the alpha wolf for hunting; when |A| > 1, the wolf pack tends to disperse to search for prey. The parameter A changes with the magnitude of the convergence factor a. The GWO algorithm is improved by updating the weight of the position, and the problem that the balance between global exploration and local development may be poor when the GWO algorithm changes linearly is integrated;

[0061] S105. Introduce a non-linear double convergence factor. In the GWO algorithm, the control algorithm for the first half of the iteration is used for global survey of the optimal value, and the second half of the iteration is used for local search of the optimal value. Although the convergence factor a changes linearly, this may result in a situation where the global and local optimization cannot reach a balance. To solve this problem, THW-GWO proposes a new method. First, calculate the average fitness value of the wolf pack. Wolves with a fitness value higher than the average are hunting wolves, while those lower than the average are scouting wolves. Corresponding non-linear double convergence factors are determined according to the classification. The improvement of the convergence factor ensures that the convergence factor a1 decreases slowly in the early and late stages and rapidly in the middle stage, achieving a balance between global search and local development, thereby improving the convergence speed of the algorithm, and making the convergence factor a2 decrease non-linearly to 0, enhancing the global exploration ability of the scouting wolves;

[0062] S106. Introduce the global alpha wolf γ to update the global optimal value. The GWO algorithm is improved by introducing an adaptive weight coefficient and the Levy flight strategy, and the problem that the balance between global exploration and local development may be poor when the GWO algorithm changes linearly is integrated. Denote the global alpha wolf γ as the global optimal value in the iteration process. The work of the γ wolf is to communicate with the hunting wolves, update the global optimal solution, and guide the scouting wolves to search for prey. Denote the position of the γ wolf as Pbest, update the position, the Levy flight is used in the update formula, and in order to improve the search efficiency of the scouting wolves, when a poor position is obtained after its Levy flight, the position is not changed, and the position of the scouting wolf is updated again.

[0063] Preferably, the step S104 specifically includes:

[0064] Adopt a new jumping-out mechanism to avoid the dilemma caused by the misjudgment of the alpha wolf leading the hunting in the GWO algorithm. When the position of the alpha wolf remains unchanged after continuous iteration for n times, let the wolf pack perform a Levy flight to continue searching for other optimal solutions until the algorithm ends.

[0065] Preferably, the step S106 specifically includes:

[0066] Introduce the adaptive weight coefficient of the double-headed wolf mechanism. The GWO algorithm uses the average weight update coefficient, but this strategy will greatly slow down the hunting speed. The THW-GWO algorithm introduces the double-headed wolf mechanism, which is responsible for the hunting wolf and the scouting wolf respectively. Since the goal of the hunting wolf is to surround and kill the prey, while the task of the scouting wolf is to search for the prey. Therefore, the THW-GWO algorithm proposes a strategy that the moving weight of the hunting wolf is determined by the decreasing speed of the fitness value. This strategy avoids the concentration of gray wolf individuals towards the α wolf, balances the influence among the α wolf, β wolf, and δ wolf, thereby enhancing the exploration ability of the algorithm.

[0067] Preferably, the step S2 specifically includes:

[0068] Condition 1, set under uniform light conditions, a light mutation occurs in the entire photovoltaic array at 2 s, that is: the solar irradiance is uniformly 1000 W / m within 0 - 2 s 2 , and the solar irradiance is uniformly 600 W / m within 2 - 4 s 2 .

[0069] Condition 2, set under partial shading conditions, the solar irradiance of each photovoltaic single cell is 1000 W / m within 0 - 2 s 2 , 900 W / m 2 , 800 W / m 2 , 700 W / m 2 , 600 W / m 2 , and the solar irradiance of each photovoltaic single cell is 800 W / m within 2 - 4 s 2 , 600 W / m 2 , 400 W / m 2 , 300 W / m 2 , 200 W / m 2 .

[0070] Simulation information setting: When simulating, four aspects of photovoltaic output power, output current, output voltage, and photoelectric conversion efficiency are selected for analysis. The MPPT tracking efficiency is defined as the ratio of the maximum output power actually tracked by the photovoltaic system to the maximum output power that can be achieved theoretically under the same environment (solar irradiance and temperature) conditions. Build a simulation model of the photovoltaic power generation system in Matlab / Simulink, and the photovoltaic array adopts a 5×1 series structure. Two simulation conditions are set during the simulation, and the temperature is always maintained at 25 °C. The control effects of the photovoltaic array under four MPPT methods of P&O, GWO, THW-GWO, and THW-GWO-P&O are compared and analyzed, and the specific applications of each algorithm are realized through the S-function module.

[0071] Preferably, the step S5 specifically includes:

[0072] S501. Initialize the algorithm, and initialize the positions of 10 wolves. These positions are evenly distributed between 0.1 and 1, representing the duty cycle of the photovoltaic array.

[0073] S502. Evaluate the fitness of the algorithm. Select the output power P as the fitness function of the photovoltaic array, and evaluate the performance of each wolf. Select the top three wolves in terms of fitness, and use their position information to guide other grey wolves.

[0074] S503. Update the positions of the algorithm. Update the positions of the wolves according to a specific formula to make them approach the maximum power point. In this stage, utilize the global search performance of the THW - GWO algorithm to enable the wolf pack to quickly converge near the maximum power point.

[0075] S504. Conduct local search for the algorithm. When the maximum number of iterations is reached or close to the MPP, adopt the P&O algorithm with a small step size for local search. The P&O algorithm has good local search performance and fast convergence performance, and can achieve precise adjustment near the maximum power point. In this stage, use the P&O algorithm with a small step size for local search until the most accurate MPP position is determined. When the termination condition is met, output the optimal value and the optimal duty cycle.

[0076] Combined with Figure 1 Figure 5 , the present invention takes the different solar irradiances and temperatures of each single photovoltaic cell of the photovoltaic array as data inputs, and takes voltage, current, power, and efficiency as data outputs. Run the simulation model of the photovoltaic power generation system, observe the gap between the actual voltage, current, power and the ideal voltage, current, power of the THW - GWO - P&O algorithm under different conditions, and conduct a comparative analysis of the efficiencies of different algorithms under the same and different conditions. The constructed simulation model respectively simulates the global maximum power point tracking under uniform illumination conditions and shading conditions. Through comparative analysis with the P&O algorithm, GWO algorithm, and THW - GWO algorithm, verify the accuracy and speed of the THW - GWO - P&O algorithm for tracking the maximum power point under complex illumination conditions. It adjusts the traditional grey wolf algorithm GWO based on swarm intelligence in terms of its convergence factor, weighted distance, position update weight, and jumping out of local optimum, and combines it with the perturbation observation method P&O, thereby adjusting the strategy of the maximum power point tracking algorithm to enable the energy storage inverter to operate at the maximum power point to the greatest extent and improve the efficiency of the energy storage inverter; in addition, the present invention integrates the advantages of different MPPT algorithms and has more advantages in terms of dynamic response and stable accuracy compared with a single MPPT control method.

[0077] Example 2

[0078] Refer to Figure 6, this embodiment relates to an improved grey wolf MPPT control device for a photovoltaic array applicable to energy storage power station charging, including a memory and one or more processors. Executable code is stored in the memory. When the one or more processors execute the executable code, it is used to implement the improved grey wolf MPPT control method for a photovoltaic array applicable to energy storage power station charging in Embodiment 1.

[0079] Embodiment 3

[0080] This embodiment relates to a computer-readable storage medium with a program stored thereon. When the program is executed by a processor, it implements the improved grey wolf MPPT control method for a photovoltaic array applicable to energy storage power station charging in Embodiment 1.

[0081] The above-disclosed are only the preferred embodiments of the present invention. Of course, the scope of rights of the present invention cannot be limited thereby. Therefore, equivalent changes made according to the scope of the patent application of the present invention still fall within the scope covered by the present invention.

Claims

1. An improved Gray Wolf MPPT control method for photovoltaic arrays used for charging energy storage power stations, characterized in that: The improved Grey Wolf algorithm THW-GWO-P&O is used to adjust the duty cycle of the Boost circuit through the MPPT controller to keep the output power of the photovoltaic array near the maximum power point. The input of the photovoltaic array is the solar irradiance and temperature, and the output is the voltage, current, power, and efficiency, including the following steps: S1. Construct an improved grey wolf algorithm THW-GWO-P&O; combine the traditional grey wolf algorithm GWO with maximum power tracking, improve the GWO algorithm, improve the convergence factor, and adjust the weighted distance and position update weight to form an improved grey wolf algorithm THW-GWO led by two wolves, and finally combine it with the traditional P&O algorithm to form an improved grey wolf algorithm THW-GWO-P&O; S2. Set simulation conditions and design the operating data of the photovoltaic array under different conditions, wherein the conditions are divided into uniform illumination and partial shading conditions, and the operating data are different solar irradiance and temperature of each single photovoltaic cell in the photovoltaic array; S3. A simulation model of a photovoltaic power generation system is built in Matlab / Simulink, including a photovoltaic array, an MPPT controller, a Boost circuit and a load component. The photovoltaic array is composed of multiple photovoltaic cells, and the circuit duty cycle is controlled by the MPPT controller using the THW-GWO-P&O algorithm. S4. Using different solar irradiance and temperature of each single photovoltaic cell in the photovoltaic array as data input, and using voltage, current, power and efficiency as data output, run the simulation model of the photovoltaic power generation system, observe the gap between the actual voltage, current and power of the THW-GWO-P&O algorithm under different conditions and the ideal voltage, current and power, and compare and analyze the efficiency of different algorithms under the same and different conditions; S5. The constructed simulation model simulates the global maximum power point tracking under uniform illumination conditions and shade conditions respectively. By comparing and analyzing with the P&O algorithm, GWO algorithm and THW-GWO algorithm, the accuracy and speed of the THW-GWO-P&O algorithm in tracking the maximum power point under complex illumination conditions are verified.

2. The improved Gray Wolf MPPT control method for a photovoltaic array suitable for charging an energy storage power station as claimed in claim 1, characterized in that: The step S1 of combining the GWO algorithm with the maximum power tracking includes: S101, GWO algorithm to find the optimal solution. The GWO algorithm is based on the principles of search, encirclement and hunting. The gray wolf continuously iterates by adjusting its position, gathers in the direction of the optimal solution, and finally achieves the goal; S102, formulate the optimization rules of the GWO algorithm, and after calculating the fitness of the gray wolf each time, compare it among the α wolf, β wolf and δ wolf to determine the new leader; the updating process is as follows: first, compare the average fitness of the gray wolf with that of the α wolf, if it exceeds, it becomes the new leader, otherwise it remains unchanged; then, compare the fitness with that of the β wolf, if it exceeds, it will be replaced; if the fitness is not as good as that of the α wolf and the β wolf, compare it with the δ wolf, if it exceeds, it will become the new δ wolf; if the fitness of the gray wolf is lower than that of the α wolf, the β wolf and the δ wolf, the leadership level remains unchanged; S103, obtaining the global maximum power, using the real-time power of the photovoltaic array as the fitness function, and the position of the gray wolf corresponds to the duty cycle; with each round of updates, the gray wolf group gradually approaches the global maximum power point; by setting different photovoltaic array input data to obtain the operation data of the GWO algorithm under different working conditions, in this process, the fitness of wolf α is considered to be the maximum output power.

3. The improved Gray Wolf MPPT control method for a photovoltaic array suitable for charging an energy storage power station as claimed in claim 2, characterized in that: The step S103 of obtaining the operating data of the GWO algorithm under different working conditions specifically includes: S1031, calculating the efficiency of the GWO algorithm, under the current uniform and partial shade lighting conditions, recording the difference between the actual voltage, current, and power of the GWO algorithm under different conditions and the ideal voltage, current, and power, and judging the efficiency; S1032. Extract the variation characteristics of the GWO algorithm, study the GWO algorithm under the condition that the solar photovoltaic cell is in partial shading, track the unique global maximum power point existing in multiple peaks of the characteristic curve, and extract the variation characteristics between the ideal power and the actual power.

4. The improved Gray Wolf MPPT control method for a photovoltaic array suitable for charging an energy storage power station as claimed in claim 1, characterized in that: The improved convergence factor described in step S1 specifically includes: S104, improve the convergence factor and weighted distance; the global and local search capabilities of the GWO algorithm are affected by parameter A; when | A |≤1, the wolf pack tends to follow the leader to hunt; when | A |>1, the wolf pack tends to disperse to find prey. Parameter A changes with the size of the convergence factor a. The GWO algorithm is improved by updating the weights through the position, and the problem of poor balance between global exploration and local development may be caused by the integration of the GWO algorithm when it changes linearly; S105, introduce nonlinear dual convergence factor. In the GWO algorithm, the control algorithm of the first half of the iteration is used to globally survey the optimal value, and the second half of the iteration is used to locally search for the optimal value. Although the convergence factor a changes linearly, this may lead to a situation where the global and local optimization cannot reach a balance. In order to solve this problem, THW-GWO proposes a new method. First, the average fitness value of the wolf pack is calculated. The wolves with a fitness value higher than the average are hunting wolves, while the wolves with a fitness value lower than the average are scout wolves. The corresponding nonlinear dual convergence factor is determined according to the classification. The improvement of the convergence factor ensures that the convergence factor a1 decreases slowly in the early and late stages and decreases rapidly in the middle stage, achieving a balance between global search and local development, thereby improving the convergence speed of the algorithm, and making the convergence factor a2 decrease from nonlinearity to 0, enhancing the global exploration ability of the scout wolf. S106, introduce the global leader wolf γ to update the global optimal value, improve the GWO algorithm, and integrate the problem that the GWO algorithm may cause poor balance between global exploration and local development when it changes linearly by introducing adaptive weight coefficients and Levy flight strategies; the global leader wolf γ is denoted as the global optimal value in the iterative process. The work content of the γ wolf is to communicate with the hunting wolf, update the global optimal solution, and guide the scout wolf to search for prey; the position of the γ wolf is denoted as P best , the position is updated, and Levy flight is used in the update formula. In order to improve the search efficiency of the scout wolf, if a bad position is obtained after the Levy flight, the position of the scout wolf is not changed, and the position is updated again.

5. The improved Gray Wolf MPPT control method for a photovoltaic array suitable for charging an energy storage power station as claimed in claim 4, characterized in that: The updating of the weight by location in step S104 specifically includes: A new jump-out mechanism is adopted to avoid the dilemma caused by the misjudgment of the α wolf leading the hunt in the GWO algorithm; when the position of the α wolf remains unchanged after n consecutive iterations, the wolf pack is allowed to perform a Levy flight to continue searching for other optimal solutions until the algorithm ends.

6. The improved Gray Wolf MPPT control method for a photovoltaic array suitable for charging an energy storage power station as claimed in claim 4, characterized in that: The step S106 of introducing the adaptive weight coefficient includes: The adaptive weight coefficient of the two-headed wolf mechanism is introduced. The GWO algorithm uses the average weight update coefficient, but this strategy will greatly slow down the hunting speed. The THW-GWO algorithm introduces the two-headed wolf mechanism, which is responsible for hunting wolves and scouting wolves respectively. Since the goal of hunting wolves is to surround and kill prey, and the task of scouting wolves is to search for prey, the THW-GWO algorithm proposes a strategy in which the movement weight of hunting wolves is determined by the speed of decreasing fitness value. This strategy avoids the gray wolf individuals from concentrating on α wolf, balances the influence between α wolf, β wolf and δ wolf, and thus enhances the exploration ability of the algorithm.

7. The improved Gray Wolf MPPT control method for a photovoltaic array suitable for charging an energy storage power station as claimed in claim 1, characterized in that: The setting of the simulation working condition described in step S2 specifically includes: Working condition 1, set under uniform lighting conditions; Working condition 2: set under partial shading conditions; Simulation information setting: photovoltaic output power, output current, output voltage and photoelectric conversion efficiency are selected for analysis during simulation; MPPT tracking efficiency is defined as the ratio of the maximum output power actually tracked by the photovoltaic system to the maximum output power that can be achieved theoretically under the same environmental conditions; a simulation model of the photovoltaic power generation system is built in Matlab / Simulink, and the photovoltaic array adopts a 5×1 series structure; two simulation conditions are set during simulation, and the temperature is always maintained at 25°C. The control effects of the photovoltaic array under the four MPPT methods of P&O, GWO, THW-GWO, and THW-GWO-P&O are compared and analyzed, and the specific application of each algorithm is realized through the S-function module.

8. The improved Gray Wolf MPPT control method for a photovoltaic array suitable for charging an energy storage power station as claimed in claim 1, characterized in that: The simulation model constructed in step S5 simulates the global maximum power point tracking under uniform illumination conditions and shade conditions, respectively, and verifies the accuracy and speed of the THW-GWO-P&O algorithm in tracking the maximum power point under complex illumination conditions by comparing and analyzing with the P&O, GWO and THW-GWO algorithms, specifically including: S501, initializing the algorithm, initializing the positions of 10 wolves, and these positions are evenly distributed between 0.1 and 1, indicating the duty cycle of the photovoltaic array; S502, evaluating the fitness of the algorithm, selecting the output power P as the fitness function of the photovoltaic array, and evaluating the performance of each wolf; selecting the top three wolves in terms of fitness, and using their position information to guide other gray wolves; S503, updating the position of the algorithm, and updating the position of the wolves according to a specific formula to make them approach the maximum power point; in this stage, the global search performance of the THW-GWO algorithm is used to make the wolf pack quickly converge to the vicinity of the maximum power point; S504. Perform local search on the algorithm. When the maximum number of iterations is reached or the MPP is approached, a small-step P&O algorithm is used for local search. The P&O algorithm has good local search performance and fast convergence performance, and can achieve precise adjustment near the maximum power point. In this stage, a small-step P&O algorithm is used for local search until the most accurate MPP position is determined. When the termination condition is met, the optimal value and the optimal duty cycle are output.

9. An improved Gray Wolf MPPT control device for photovoltaic arrays used for charging energy storage power stations, characterized in that: It comprises a memory and one or more processors, wherein the memory stores executable codes, and when the one or more processors execute the executable codes, they are used to implement the improved Grey Wolf MPPT control method for a photovoltaic array suitable for charging an energy storage power station as described in any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that: A program is stored thereon, and when the program is executed by a processor, the improved Grey Wolf MPPT control method for a photovoltaic array suitable for charging an energy storage power station as described in any one of claims 1 to 8 is implemented.

Images