Photovoltaic cell maximum power point tracking method based on improved sparrow search algorithm

Abstract

The invention provides a photovoltaic cell maximum power point tracking method based on an improved sparrow search algorithm, relates to the technical field of photovoltaic power generation, and solves the problems that a traditional photovoltaic cell maximum power point tracking algorithm still falls into a local extreme value in a local shadow environment, so that the tracking precision is sharply reduced, and the tracking time is shortened. And the problem of failure in a local shadow environment is solved. According to the technical scheme, the method comprises the steps that a composite algorithm of an improved sparrow search algorithm and a variable step size perturbation and observation method is used as a control algorithm of an MPPT controller, iteration is conducted according to the flow of the improved sparrow search algorithm after the improved sparrow search algorithm is initialized at the initial stage of algorithm operation, and the output power of a photovoltaic cell array under the current duty ratio is calculated; the output power is the fitness value of the algorithm, and the global optimal position is updated. The method can adapt to a rapidly changing illumination environment, power oscillation is obviously reduced in the process of tracking the maximum power point, stable output of a photovoltaic cell can be guaranteed, and stable operation of a communication base station is guaranteed.

Description

Technical Field

[0001] The present invention relates to the field of photovoltaic power generation technology, in particular to a photovoltaic cell maximum power point tracking method based on an improved sparrow search algorithm, which is suitable for efficient power generation control of photovoltaic cells in a local shadow environment. Background Art

[0002] With the widespread adoption of 5G communication technology, the energy consumption of communication base stations has become increasingly prominent. Photovoltaic power generation, as a clean energy source, has been widely used in communication base station equipment. Photovoltaic cells are a key component of solar power generation, directly converting solar energy into electricity and supplying it to terminal devices. However, the output power of photovoltaic cells is significantly affected by environmental factors such as light intensity and temperature, and localized shadows can cause multiple peaks in the output characteristic curve. Research on maximum power point tracking technology allows for real-time monitoring of the output power of photovoltaic cell arrays and timely adjustment of the operating state of photovoltaic cells based on the environment to ensure maximum power output under varying environmental conditions.

[0003] Traditional maximum power point tracking (MPPT) algorithms, including perturbation and observation and conductance increment methods, have limitations. The existing perturbation and observation (P&O) method uses periodic voltage perturbations to find the maximum power point, but its fixed step size makes it difficult to balance tracking speed and accuracy. It is prone to misjudgment during sudden changes in illumination and prone to falling into local minima under partial shadows, resulting in reduced power generation efficiency. The conductance increment method, based on slope comparison, offers high theoretical accuracy but is computationally complex, has demanding sensor requirements, and high hardware costs. Existing intelligent algorithms, including the particle swarm optimization (PSO) and the sparrow search algorithm, also exhibit shortcomings in practical applications. The particle swarm optimization (PSO) algorithm is prone to premature convergence and lacks global search capabilities under partial shadows. The sparrow search algorithm (SSA) algorithm has poor initial population randomness, and its boundary handling strategy reduces diversity. It is prone to falling into local optima under multimodal functions. The algorithm's performance is significantly affected by the initial solution, and a poor initial solution can cause the algorithm to fall into local minima and slow its convergence. Therefore, there is an urgent need for an MPPT method that takes into account both global search capability and local optimization accuracy to improve the stability and efficiency of photovoltaic systems in complex environments. Summary of the Invention

[0004] The present invention aims to provide a photovoltaic cell maximum power point tracking method based on an improved sparrow search algorithm. This method addresses the problem that conventional photovoltaic cell maximum power point tracking algorithms can still fall into local extremes in partially shaded environments, resulting in a sharp drop in tracking accuracy and failure in partially shaded environments. The method can track the maximum power point quickly and accurately in partially shaded environments.

[0005] The technical solution adopted by the present invention is: the photovoltaic cell maximum power point tracking method based on the improved sparrow search algorithm includes: using a composite algorithm of the improved sparrow search algorithm and the variable step size perturbation observation method as the control algorithm of the MPPT controller, the steps are: in the early stage of the algorithm operation, first using the improved sparrow search algorithm, after completing the initialization of the algorithm parameters, iterating according to the process of the improved sparrow search algorithm, adjusting the duty cycle output of the circuit, and calculating the output power of the photovoltaic cell array under the current duty cycle, the output power is the fitness value of the algorithm, and the global optimal position is updated; when the improved sparrow search algorithm iterates a set number of times, the algorithm After fully exploring the entire search space and converging to the vicinity of the global maximum power point, it can be regarded as the maximum power point tracking of the photovoltaic array under uniform illumination conditions. At this time, the improved perturbation observation algorithm is switched to quickly track the maximum power point. The position of the optimal sparrow individual in the improved sparrow search algorithm before the algorithm switch is the initial position for the variable step size perturbation observation algorithm to search for the optimal value. When the light intensity of the working environment of the photovoltaic array changes and the output power changes significantly, it is necessary to restart the composite algorithm of the improved sparrow search algorithm and the variable step size perturbation observation method to track the new maximum power point again. The restart condition formula (1): |PP old |>0.05×P old In formula (1), P represents the current power value, P old Indicates the maximum power value recorded last time;

[0006] Among them, the specific process of improving the sparrow search algorithm is as follows:

[0007] S1. Set the initial parameters of the sparrow search algorithm: population size, search space boundary, safety value and warning value;

[0008] S2. The chaotic map completes population initialization and divides the sparrow population into discoverers, joiners, and guards according to the set ratio;

[0009] S3. Calculate the fitness value of each individual according to the corresponding functional relationship;

[0010] S4. According to formula (2)

[0011] Formula (3)

[0012] Formula (4) Complete the location update of the discoverer, joiner and vigilant;

[0013] In formula (2), t represents the current number of iterations, t max represents the maximum number of iterations; j represents the dimension of the variable to be optimized; X (n,j)represents the position of the nth sparrow in the jth dimension; c is a random number between 0 and 1, and the value of c is not 0; R2 is the warning value of the sparrow population, usually in the range of R2∈[0,1]; S T It is a pre-set safety value, usually S T ∈[0.5,1]; Q is a random number that obeys the normal distribution; A is a matrix whose elements are all 1, the number of rows is 1, and the number of columns is the same as the dimension of the variable; at the same time, construct a matrix B, which has the same number of rows and columns as A, and the elements in the matrix B are randomly selected between -1 and 1, and B + =B T (BB T ) -1 ;

[0014] In formula (3), N is the number of sparrow population; X w,t is the position of the individual with the lowest energy in the current sparrow population, X P,t is the position of the individual with the highest energy among the current discoverers;

[0015] In formula (4), X b,t is the position with the highest energy in the current sparrow population; b is a random number that follows a normal distribution with a mean of 0 and a variance of 1, which is used to control the step size of the position update of the sentinel; c1 is a random number that represents the direction of the sparrow's movement and controls the step size of the position update, c1∈[-1,1]; f n is the fitness value of the current sparrow individual; f g and f w are the fitness values ​​of the current global best position and the worst position respectively;

[0016] S5. For individuals that exceed the search boundary, they are processed according to the proposed improved boundary processing strategy to bring them back into the search space.

[0017] S6. Using the reverse learning strategy according to formula (5) X2 = l + u - X1 to generate a reverse population and calculate the individual fitness value. By comparing the fitness values, the sparrow individual with the higher fitness value is selected to enter the next iteration;

[0018] S7. Determine whether the algorithm termination condition is met: If the termination condition is met, output the optimal individual and its corresponding fitness value; if the termination condition is not met, return to step S4 and continue running until the termination condition is met.

[0019] When the photovoltaic cell array works in a high temperature environment, a composite algorithm of an improved sparrow search algorithm and a variable step size perturbation observation method is used to track the maximum power point of the photovoltaic array; under medium and low temperature conditions, an improved sparrow search algorithm is used to track the maximum power point of the photovoltaic array.

[0020] When the photovoltaic cell array operates in a low temperature environment of 5°C–15°C, an improved sparrow search algorithm is used to track the maximum power point of the photovoltaic array. When the photovoltaic cell array operates in a low temperature environment of 0°C–less than 5°C, a sparrow search algorithm is used to track the maximum power point of the photovoltaic array.

[0021] The advantages and positive effects of the present invention are as follows: the photovoltaic cell maximum power point tracking method based on the improved sparrow search algorithm utilizes a composite algorithm of the improved sparrow search algorithm and the variable step-size perturbation observation method as the control algorithm of the MPPT controller. The improved sparrow search algorithm uses a chaotic map instead of the random population initialization of the sparrow search algorithm, so that the generated initial population has diversity and improves the convergence speed of the algorithm. Secondly, the reflection method is used to improve the boundary processing strategy, so that sparrow individuals outside the search space can return to the search space, ensuring the diversity of the population, enabling full exploration of the search space, and improving the global search capability of the algorithm. In each iteration, a reverse learning strategy is used to select the best to enter the next iteration, accelerating the convergence of the population to the optimal value and further improving the search speed of the algorithm. The hybrid algorithm has extremely high optimization speed and optimization accuracy, solves the problem that the variable step-size perturbation observation method IP&O algorithm fails in partial shadow environments, can adapt to rapidly changing lighting environments, and significantly reduces power oscillation during the tracking of the maximum power point, ensuring stable output of photovoltaic cells and ensuring the smooth operation of communication base stations. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The present invention will be further described below with reference to the accompanying drawings.

[0023] Figure 1 This is a control flow chart of a composite algorithm of an improved sparrow search algorithm and a variable step-size perturbation-observation method according to an embodiment of the present invention;

[0024] Figure 2 Flowchart of an improved sparrow search algorithm according to an embodiment of the present invention;

[0025] Figure 3 A schematic diagram illustrating a reverse solution generation principle of a reverse learning strategy according to an embodiment of the present invention;

[0026] Figure 4 A schematic diagram of a cropping boundary processing strategy according to an embodiment of the present invention;

[0027] Figure 5 This is a schematic diagram of the improved boundary processing principle of the reflection method processing strategy according to an embodiment of the present invention;

[0028] Figure 6 A comparison diagram of the convergence curves of the four algorithms of the embodiment of the present invention on seven functions;

[0029] Figure 7PU output characteristic curve of the photovoltaic cell array under uniform illumination according to an embodiment of the present invention;

[0030] Figure 8 This is a comparison diagram of output power characteristic curves of simulation results under uniform illumination conditions for four algorithms according to an embodiment of the present invention;

[0031] Figure 9 PU output characteristic curve diagram of a photovoltaic cell array under partial shadow according to an embodiment of the present invention;

[0032] Figure 10 This is a comparison diagram of output power characteristic curves of simulation results under partial shadow conditions of four algorithms in an embodiment of the present invention;

[0033] Figure 11 PU output characteristic curve diagram of a photovoltaic cell array under dynamic shadow according to an embodiment of the present invention;

[0034] Figure 12 This is a comparison diagram of output power characteristic curves of simulation results under dynamic shadow conditions of four algorithms in an embodiment of the present invention;

[0035] Figure 13 is the IV curve of the photovoltaic array according to an embodiment of the present invention;

[0036] Figure 14 is a PV curve of a photovoltaic array according to an embodiment of the present invention;

[0037] Figure 15 is a Pearson correlation diagram of an embodiment of the present invention;

[0038] Figure 16 This is a three-dimensional scatter plot of light intensity, temperature, and power at the maximum power point of a solar panel according to an embodiment of the present invention. DETAILED DESCRIPTION

[0039] according to Figure 1-16 Describe the specific solutions of the present invention in detail, such as Figure 1 and Figure 2 As shown in the embodiment, a photovoltaic cell maximum power point tracking method using an improved sparrow search algorithm is provided. The maximum power point tracking method is referred to as the MPPT algorithm; the sparrow search algorithm is referred to as the SSA algorithm; the improved sparrow search algorithm is referred to as the ISSA algorithm; the variable step size perturbation and observation method is referred to as the IP&O algorithm; and the composite algorithm of the improved sparrow search algorithm and the variable step size perturbation and observation method is referred to as ISSA-IP&O. Figure 1The control flow chart of the composite algorithm of the improved sparrow search algorithm and the variable step-size perturbation observation method is shown in the figure. The composite algorithm of the improved sparrow search algorithm and the variable step-size perturbation observation method is used as the control algorithm of the MPPT controller. The specific steps are: in the early stage of the algorithm operation, the improved sparrow search algorithm is first used. After the initialization of the algorithm parameters is completed, it is iterated according to the process of the improved sparrow search algorithm to adjust the duty cycle output of the circuit and calculate the output power of the photovoltaic cell array under the current duty cycle. The output power is the fitness value of the algorithm and the global optimal position is updated. When the improved sparrow search algorithm iterates the set number of times, the algorithm completes the entire The full exploration of the search space converges to the vicinity of the global maximum power point, which can be regarded as the maximum power point tracking of the photovoltaic array under uniform illumination conditions. At this time, the improved perturbation observation algorithm is switched to quickly track the maximum power point; the position of the optimal sparrow individual in the improved sparrow search algorithm before the algorithm switch is the initial position for the variable step size perturbation observation algorithm to search for the optimal value; when the light intensity of the working environment of the photovoltaic array changes and causes a significant change in the output power, it is necessary to restart the composite algorithm of the improved sparrow search algorithm and the variable step size perturbation observation method to track the new maximum power point again; the restart condition formula (1): |PP old |>0.05×P old In formula (1), P represents the current power value, P old Indicates the maximum power value recorded last time; Figure 2 The flowchart of the improved sparrow search algorithm is shown in FIG. The specific process of the improved sparrow search algorithm is as follows:

[0040] S1. Set the initial parameters of the SSA algorithm: population size, search space boundary, safety value and warning value;

[0041] S2. The chaotic map completes population initialization and divides the sparrow population into discoverers, joiners, and guards according to the set ratio;

[0042] S3. Calculate the fitness value of each individual according to the corresponding functional relationship;

[0043] S4. According to formula (2):

[0044] Formula (3):

[0045] Formula (4): Complete the location update of the discoverer, joiner and vigilant;

[0046] In formula (2), t represents the current number of iterations, t max represents the maximum number of iterations; j represents the dimension of the variable to be optimized; X (n,j)Represents the position of the nth sparrow individual in the jth dimension; c is a random number between 0 and 1 and the value of c is not 0. R2 is the warning value of the sparrow population, usually in the range of R2∈[0,1]; S T It is a pre-set safety value, usually S T ∈[0.5,1]; Q is a random number that obeys the normal distribution; A is a matrix whose elements are all 1, the number of rows is 1, and the number of columns is the same as the dimension of the variable; at the same time, construct a matrix B, which has the same number of rows and columns as A, and the elements in the matrix B are randomly selected between -1 and 1, and B + =B T (BB T ) -1 ;

[0047] In formula (3), N is the number of sparrow population; X w,t is the position of the individual with the lowest energy in the current sparrow population, X P,t is the position of the individual with the highest energy among the current discoverers.

[0048] In formula (4), X b,t is the position with the highest energy in the current sparrow population; b is a random number that follows a normal distribution with a mean of 0 and a variance of 1, which is used to control the step size of the position update of the sentinel; c1 is a random number that represents the direction of the sparrow's movement and controls the step size of the position update, c1∈[-1,1]; f n is the fitness value of the current sparrow individual; f g and f w are the fitness values ​​of the current global best position and the worst position respectively;

[0049] S5. For individuals that exceed the search boundary, they are processed according to the proposed improved boundary processing strategy to bring them back into the search space.

[0050] S6. Using the reverse learning strategy according to formula (5) X2 = l + u - X1 to generate a reverse population and calculate the individual fitness value. By comparing the fitness values, the sparrow individual with the higher fitness value is selected to enter the next iteration;

[0051] S7. Determine whether the algorithm termination condition is met: If the termination condition is met, output the optimal individual and its corresponding fitness value; if the termination condition is not met, return to step S4 and continue running until the termination condition is met.

[0052] The principle of photovoltaic cell maximum power point tracking technology: Research on the output characteristics of photovoltaic cells shows that their output characteristic curves vary significantly with changes in light intensity and ambient temperature, and maintain a fixed global maximum power point at any light intensity and ambient temperature. In real-world applications, the operating environment of photovoltaic cells changes in real time, and their maximum power point is not fixed. Therefore, in order to adjust the operating point of the photovoltaic cell array to maintain maximum power output when external light intensity or ambient temperature changes, in other words, to achieve impedance matching between the equivalent internal resistance of the photovoltaic cell array and the external load resistance, this can be achieved by changing the duty cycle.

[0053] The output power of photovoltaic cells is affected by a variety of external environmental factors, with temperature and light intensity being the two most critical parameters. While light intensity has a more direct and significant impact on output power, temperature changes can also have a profound impact on the electrical characteristics of photovoltaic cells, thereby affecting the location of the maximum power point and power output level. Specifically, the impact of temperature is mainly reflected in the following aspects:

[0054] 1. Open circuit voltage (Voc) decreases significantly with increasing temperature

[0055] The open circuit voltage of photovoltaic cells is extremely sensitive to temperature, and its variation can be described by the following empirical formula (6):

[0056]

[0057] Where V OC is the open circuit voltage, which is the output voltage of the battery when there is no load, and T represents the temperature. This formula indicates the drop in open-circuit voltage for every 1°C increase in temperature. This formula demonstrates that the open-circuit voltage of solar cells is very sensitive to temperature, with a voltage drop of approximately 2.2–2.5 mV (per cell) for every 1°C increase in temperature. Across the entire photovoltaic array, these small changes can accumulate into significant voltage fluctuations that must be compensated for in MPPT control.

[0058] As the temperature rises, the band gap of the semiconductor material decreases, the electron thermal motion increases, and the recombination rate increases, which makes V OC This change is particularly evident in series-connected photovoltaic modules. For every 1°C increase in temperature, the total voltage drop increases exponentially, significantly reducing the system's power output.

[0059] 2. Short-circuit current (Isc) increases slightly with temperature

[0060] With V OCDifferent from the short-circuit current, the short-circuit current is less sensitive to temperature and shows a slight upward trend. The main reason is that the increase in temperature increases the carrier concentration and prolongs the minority carrier lifetime, resulting in a slight increase in the photogenerated current per unit area. Its rate of change formula (7) is generally:

[0061]

[0062] I SC represents the short-circuit current, that is, the maximum current output by the solar cell when the load is zero (voltage is 0), and T represents the temperature. It indicates the percentage increase of short-circuit current for every 1°C increase in temperature.

[0063] 3. The maximum power tracking point (MPPT) moves downward overall

[0064] The maximum power point of a photovoltaic cell is the result of the combined effect of voltage and current. Since the impact of voltage drop is far greater than the contribution of current increase, when the temperature rises, such as Figure 13 The IV curve of the photovoltaic array shown and Figure 14 The inflection point of the PV curve of the photovoltaic array shown in the figure shifts to the left and downward, which is reflected in the downward shift of the MPPT. This means that the optimal operating point of the system output power will continue to change with temperature changes. The comparison of current and voltage changes is shown in Table (1):

[0065] Table (1) Comparison of the rate of change of current and voltage with temperature

[0066]

[0067] The voltage drops significantly with temperature, while the current increases less.

[0068] like Figure 15The Pearson correlation plot shown in Figure 2 shows the Pearson correlations between the five variables IMAX, VMAX, PMAX, GMAX, and TMAX, revealing the inherent connections between PV system operating parameters. The correlation matrix shows that maximum current (IMAX) exhibits a strong positive correlation with maximum power (PMAX) and maximum light intensity (GMAX) (r=0.97 and 0.98, respectively), indicating that under strong light conditions, the cell output current increases significantly, thereby driving up power output. Maximum power also exhibits strong positive correlations with light intensity (r=0.96) and temperature (r=0.89), indicating that light intensity is the dominant factor affecting PV power output, while temperature indirectly affects power by slightly increasing current through increased carrier concentration. However, high temperatures can cause a decrease in open-circuit voltage, offsetting some of the power gains. The strong correlation between maximum temperature (TMAX) and maximum current and maximum light intensity (r=0.93 and 0.94, respectively) reflects the trend of increasing ambient temperature with increasing solar radiation. In contrast, the correlation between maximum voltage (VMAX) and other variables is generally weak (r with IMAX, GMAX, and TMAX are all below 0.2), indicating that voltage is less affected by environmental factors and has high stability. This correlation analysis provides a theoretical basis for subsequent optimization of the MPPT (maximum power point tracking) algorithm, which can improve the system's tracking accuracy and response speed to the maximum power point by paying more attention to the dynamic changes in current and light.

[0069] Since the position of the MPPT is significantly affected by temperature, especially when operating outdoors or in areas with large temperature differences between day and night, such as plateaus and deserts, where temperature changes frequently and with large amplitudes, the MPPT system needs to have the following capabilities: (1) Fast tracking capability: able to respond to changes in the MPPT position in a timely manner to avoid energy loss due to response lag. (2) High-precision recognition capability: accurately judge the new MPPT position to avoid oscillation or misjudgment. (3) Anti-disturbance capability: maintain stable operation in the power fluctuations induced by temperature changes to prevent the algorithm from falling into a local optimal solution. Traditional perturbation and observation (P&O) or incremental admittance method (INC) may experience tracking delays or poor stability when the temperature changes rapidly. Therefore, the use of intelligent optimization algorithms with global search capabilities (such as ISSA, PSO, GWO, etc.) or hybrid MPPT algorithms that combine them with classic algorithms can more effectively deal with the MPPT offset problem caused by temperature.

[0070] The MPPT principle is as follows: by collecting real-time information on the output voltage and current of the photovoltaic cell array, it is calculated to determine whether the photovoltaic cell array is operating at its maximum power point in the current operating environment. If the current output power has not yet reached the maximum output power, the MPPT controller will convert the voltage and current signals collected from the photovoltaic cell array into corresponding circuit control signals based on the MPPT algorithm used to adjust the DC / DC circuit duty cycle in real time. This allows the photovoltaic cell array to adapt to changes in the operating environment and dynamically and stably operate at the maximum power point, ensuring that the output power is always the maximum power value. The process of tracking the maximum power point of the photovoltaic cell is the process of achieving dynamic resistance matching by adjusting the external load.

[0071] The variable-step perturbation-and-observation (IP&O) method is a classic, improved algorithm for maximum power point tracking (MPPT) in photovoltaic systems. It adaptively adjusts the magnitude and direction of the perturbation based on the current operating point position and power variations, balancing tracking speed and accuracy by dynamically adjusting the perturbation step size. The MPPT method based on the IP&O algorithm effectively addresses the difficulty of traditional P&O algorithms in balancing tracking speed and optimization accuracy under a fixed perturbation step size. It can quickly and accurately track the maximum power point, and significantly reduces stable power oscillations. However, when the photovoltaic array is partially shaded, the IP&O algorithm will similarly fall into a local extreme value and fail to track the maximum power point, failing to address the failure of traditional P&O algorithms in partial shadows.

[0072] The applicant verified its performance through a series of simulation experiments and found that the IP&O algorithm can effectively solve the problem of the traditional fixed-step P&O algorithm that has difficulty balancing tracking speed and tracking accuracy. It can quickly and accurately track the maximum power point under uniform lighting conditions, and the power oscillation after stable output is very small. However, the IP&O algorithm still falls into local extreme values ​​in partial shadow environments, resulting in a sharp drop in tracking accuracy, and fails to solve the problem of traditional algorithms failing in partial shadow environments. Therefore, a method is needed to achieve maximum power point tracking in partial shadow environments to ensure that the photovoltaic cell array can operate at the maximum power point in any environment and ensure the stable operation of the communication base station.

[0073] Therefore, the applicant proposed a method for tracking the maximum power point of photovoltaic cells using an improved sparrow search algorithm, using a composite algorithm of the improved sparrow search algorithm and the variable step-size perturbation observation method as the control algorithm of the MPPT controller. The sparrow search algorithm has a strong global search capability and can effectively avoid falling into local extreme values. The sparrow search algorithm is improved to address problems such as large power oscillations during the optimization process and long optimization time of the swarm intelligence algorithm. At the same time, in order to further improve the convergence speed of the algorithm, the improved IP&O algorithm is combined with the improved sparrow search algorithm. Finally, the proposed composite algorithm is simulated to verify its feasibility in tracking the maximum power point.

[0074] Chaotic mapping is a nonlinear dynamical system whose output exhibits complex and unpredictable behavior within a finite timeframe. The randomness and ergodicity of chaotic mapping can increase the randomness and unpredictability of the system, helping to address premature convergence and local optima in optimization problems.

[0075] The performance of the SSA algorithm is greatly affected by the initial solution. The initial solution is randomly generated and unevenly distributed in the search space. Using chaotic mapping to generate the initial population can make the initial solution randomly and evenly distributed in the search space, thereby improving the global search ability and convergence speed of the SSA algorithm.

[0076] Most intelligent algorithms use the Logistic Chaotic Map, which has an uneven distribution and a large distribution ratio in the range of [0, 0.1] and [0.9, 1]. For example, the Tent Chaotic Map, which is uniformly distributed in [0, 1], completes the population initialization. The mathematical expression of the Tent Chaotic Map is:

[0077]

[0078] In formula (8), θ∈(0,1), and the closer θ is to 0.5, the more uniform the distribution of the Tent chaotic map. However, when θ=0.5, the system will show a short-period state, which is not a chaotic system. Therefore, θ=0.5 is selected and the Tent chaotic map is improved to avoid the system showing a short-period state. The mathematical expression of the improved Tent chaotic map is:

[0079]

[0080] In formula (9), n is the number of the entire population. First, randomly generate an initial value in the range of 0 to 1, and then iterate according to formula (9) to generate the entire population. The improved population initialization formula is as follows:

[0081] X n =z i ·(ul)+l,n=1,2,…,num (10)

[0082] The Tent chaotic map is used to complete the population initialization of the SSA algorithm, and the diversity of the initial population is improved, making the SSA algorithm less likely to fall into local extreme values, improving the global search ability of the algorithm, and improving the convergence speed and stability of the algorithm.

[0083] The reverse learning strategy, first proposed by Tizhoosh, is a strategy used in optimization algorithms to enhance search capabilities and prevent premature convergence. The core idea is to generate a set of new reverse solutions in the opposite direction of the current solution. These new reverse solutions have a 50% probability of being better than the current solution and closer to the global optimal solution. After merging the current solution with the new reverse solutions, their fitness values ​​are compared using a corresponding function. Solutions with high fitness are selected to form the solution population, enhancing population diversity and fully exploring the search space.

[0084] According to the principle of reverse learning strategy, the schematic diagram of the generation principle of reverse solution is as follows Figure 3 As shown in the figure, for a given solution X1, within the set search space, the calculation formula for its reverse solution X2 is as follows:

[0085] X2=l+u-X1 (5)

[0086] The reverse solution X2 generated according to formula (5) is at the same distance from the midpoint of the search space as the original solution X1, and the reverse solution and the original solution each have a 50% probability of being closer to the optimal solution in the search space.

[0087] During each iteration of the SSA algorithm, the reverse individual is generated for each individual sparrow after each iteration according to formula (5). The fitness values ​​of these individuals are calculated and compared using the corresponding function. The sparrow individuals with higher fitness values ​​are selected to enter the next iteration, while the individuals with lower fitness values ​​are eliminated. The reverse learning strategy increases the diversity of the sparrow population during the search process, explores the search space more widely, and accelerates the SSA algorithm's search for the global optimal solution by generating the best reverse solution and selecting it, thus improving the convergence ability of the SSA algorithm.

[0088] In the widespread application of swarm intelligence algorithms, the algorithm's optimization capabilities and convergence speed are often limited by the definition of the search space. When the solution value exceeds the set search space boundary during the iteration process, using an appropriate boundary handling strategy can bring individuals that have exceeded the search space boundary back into the search space, thereby increasing the diversity of the population, fully exploring the search space, and improving the algorithm's performance.

[0089] The most commonly used boundary processing strategy is clipping. Clipping completes boundary processing in the optimization process by clipping individuals that exceed the boundary to the closest boundary. For example, if an individual exceeds the upper boundary, it will be set to the upper boundary value; if the individual is below the lower boundary, it will be set to the lower boundary value. The principle of completing boundary processing through clipping is as follows: Figure 4 As shown in the figure, when the position of an individual X in the population exceeds the search space, the position of the individual is set to (X) through the clipping method, and the position of X in the population is replaced by the position of (X) for iteration to complete the subsequent search.

[0090] When using the clipping method to process individuals that exceed the boundary, if more individuals exceed the upper boundary or the lower boundary during one iteration, these individuals will all be fixed at the upper boundary or the lower boundary, resulting in more identical values ​​that are not necessarily optimal, reducing the diversity of the population and not fully exploring the entire search space. Therefore, a more appropriate boundary processing strategy is needed to enable individuals that exceed the boundary to return to the search space and ensure the diversity of the population.

[0091] In each iteration, all individuals that exceed the boundary of the search space are reflected at the boundary to return their positions to the search space. The improved boundary processing principle diagram is shown in the figure below. Figure 5 As shown in the figure, when the position of an individual X in the population exceeds the search space, the individual is reflected at the boundary to return to (X) in the search space, and the position of (X) replaces the position of X in the population for iteration to complete the subsequent search, and the distance between (X) and X and the boundary is the same.

[0092] The process design of the improved sparrow search algorithm is used to improve the SSA algorithm by using three methods: chaotic mapping, reverse learning strategy and improved boundary processing strategy. A multi-strategy improved sparrow search algorithm (ISSA) is proposed. The initial population of the improved algorithm has greater randomness and ergodicity, which improves the diversity of the initial population. The optimal solution can be searched more quickly during the optimization process, and the values ​​that exceed the boundary during each iteration can be returned to the search space, ensuring the diversity of the population. The search space can be fully explored, and the convergence speed and global search ability of the algorithm are improved. The specific process of the ISSA algorithm is as follows Figure 2 shown.

[0093] To verify the performance of the improved sparrow search algorithm, seven standard test functions were used. Four of these standard test functions are unimodal, with a single global optimum within the search space, and are used to simulate the output of a photovoltaic array under uniform shading. The other three standard test functions are multimodal, with a single global optimum and multiple local optimums within the search space, and are used to simulate the output of a photovoltaic array under non-uniform shading. Under the same experimental conditions, the PSO algorithm, TSO algorithm, sparrow search algorithm (SSA), and improved sparrow search algorithm (ISSA) were tested. If the ISSA algorithm performs well on the standard test functions, it proves that the ISSA algorithm also has good performance in handling the maximum power point tracking problem of photovoltaic arrays.

[0094] This paper selects 7 standard test functions, and the dimension is set to 30. The specific information of the selected standard test functions is shown in Table (2).

[0095] Table (2) Standard test functions

[0096]

[0097] To ensure the fairness of the experiment, the population size of the particle swarm optimization algorithm PSO algorithm, tuna swarm optimization algorithm TSO algorithm, SSA algorithm and ISSA algorithm is set to 50, and the maximum number of iterations is 500. The performance of the above algorithms is verified using the selected 7 standard test functions. The algorithms are run 30 times independently and the optimal fitness value, the worst fitness value, the average fitness value and the standard deviation of each algorithm are recorded. These four sets of parameters are used as algorithm performance evaluation parameters for comparative analysis. The tracking results of the four algorithms under the standard test function are shown in the figure below. Figure 6 As shown in Table (3), the parameter calculation results after the four algorithms are run 30 times are shown in Table (3).

[0098] according to Figure 6 As can be seen from the experimental results in Table (3), in the single-peak test curves F1 to F4, the ISSA algorithm can always converge to the global optimal value most quickly and accurately, and both the optimization accuracy and optimization speed are improved. In the multi-peak test functions F5 to F7, the ISSA algorithm can avoid converging to the local optimal value, and the convergence speed is improved compared with the other three algorithms. Therefore, this algorithm can be applied to the MPPT control of photovoltaic cells in communication base stations to improve power generation efficiency and ensure the stable operation of communication base stations.

[0099] Table (3) Performance of four algorithms

[0100]

[0101]

[0102] The application of ISSA-IP&O algorithm in MPPT. According to simulation experiments, the improved ISSA algorithm can explore the entire search space more fully than the traditional SSA algorithm, avoid falling into local extreme values ​​during the optimization process, improve the global search capability, and improve the convergence speed; while the IP&O algorithm can quickly search for the maximum power point under uniform illumination conditions, and the steady-state power oscillation is very small. Therefore, this paper proposes a hybrid control algorithm based on the ISSA algorithm and the IP&O algorithm, namely the ISSA-IP&O algorithm, to give full play to the respective advantages of the two algorithms and quickly and accurately track the maximum power point.

[0103] When the ISSA-IP&O algorithm is used as the control algorithm of the MPPT controller, the input parameters of the MPPT controller are the output voltage and output current of the photovoltaic cell array, and the output parameter is the duty cycle D of the Boost circuit. Impedance matching is achieved by adjusting the duty cycle, thereby adjusting the output of the photovoltaic cell array so that the photovoltaic cell array operates at the maximum power point.

[0104] Based on the composite algorithm of the improved sparrow search algorithm and the variable step-size perturbation observation method, the flowchart of the ISSA-IP&O algorithm is as follows: Figure 1 shown.

[0105] Simulation Analysis,In order to verify the performance of the ISSA-IP&O algorithm in tracking the maximum power point of photovoltaic cells,,the TSO, SSA, ISSA and ISSA-IP&O algorithms are simulated and analyzed,respectively.,Experimental Simulation Model Structure and Boost Circuit Parameter Setting,After building the simulation model, set up the following three environments for,simulation experiments:

[0106] (1) Uniform illumination environment: The illumination intensity of all photovoltaic cell arrays is G = 1000W / m 2 , the ambient temperature is constant at T = 25℃.

[0107] (2) Partial shadow environment: There are two photovoltaic arrays with a constant illumination intensity of G = 1000 W / m 2 , the illumination intensity of a photovoltaic array is G = 400W / m 2 , the illumination intensity of the remaining two photovoltaic arrays is G = 600W / m 2 , the ambient temperature is constant at T = 25℃.

[0108] (3) Dynamic shadow environment: At the start time, there are two photovoltaic arrays with a light intensity of G = 1000W / m 2 , the illumination intensity of a photovoltaic array is G = 400W / m 2 , the illumination intensity of the remaining two photovoltaic arrays is G = 600W / m 2At 0.4s, the illumination intensity of the photovoltaic array changes, and the illumination intensity of the two photovoltaic arrays changes from G = 1000W / m 2 Reduced to G = 900W / m 2 ; The illumination intensity of a photovoltaic array is G = 400W / m 2 Increase to G = 500W / m 2 The illumination intensity of the remaining two photovoltaic arrays is G = 600W / m 2 Increase to G = 800W / m 2 During the experiment, the ambient temperature was kept constant at T = 25°C.

[0109] The PU output characteristic curve of the photovoltaic cell array under uniform illumination conditions is as follows: Figure 7 As shown in the figure, under uniform illumination conditions, the PU output characteristic curve of the photovoltaic cell array is a single-peak curve with a single maximum power point. The theoretical maximum output power is P m =8517.8W.

[0110] Under uniform illumination conditions, the MPPT performance of the TSO algorithm, SSA algorithm, ISSA algorithm and ISSA-IP&O algorithm are simulated respectively, and the output power characteristic curves are compared. Figure 8 As shown in the figure, under uniform illumination, the four algorithms can all track the maximum power point and ensure stable power output. The maximum output powers are 8469.65W, 8517.77W, 8517.46W and 8517.5W respectively, and the tracking accuracy is 99.43%, 99.98%, 99.99% and 99.99% respectively. From the above experimental results, it can be seen that under uniform illumination, the four algorithms can all track the maximum power point and ensure that the photovoltaic cell array operates stably at the maximum power point. The steady-state output power is very close to the theoretical maximum output power, and a relatively accurate maximum power point tracking control can be achieved. The comparison of simulation experimental results is shown in Table (4).

[0111] Table (4) Tracking performance under uniform lighting conditions

[0112]

[0113] As can be seen from the data in Table (4), the TSO algorithm tracks the maximum power point and maintains stable output in 0.298s, the SSA algorithm tracks the maximum power point and maintains stable output in 0.311s, the ISSA algorithm tracks the maximum power point and maintains stable output in 0.262s, and the ISSA-IP&O algorithm tracks the maximum power point and maintains stable output in 0.194s. Compared with the above tracking times, the ISSA-IP&O algorithm shortens the tracking time by 34.90%, 37.62%, and 25.95%, respectively.

[0114] Comparing the results of maximum power point tracking using the IP&O algorithm, it can be seen that the tracking accuracy of the IP&O algorithm and the ISSA-IP&O algorithm is not much different. However, the IP&O algorithm only takes 0.04s to track the maximum power point, which is 79.38% shorter than the ISSA-IP&O algorithm. It is more suitable for tracking the maximum power point in a uniform lighting environment.

[0115] The PU output characteristic curve of the photovoltaic cell array under partial shadow environment is as follows: Figure 9 As shown in the figure, under partial shadow environment, the PU output characteristic curve of the photovoltaic cell array presents a multi-peak characteristic with three peaks, which are the theoretical maximum output power P of the photovoltaic cell array and the theoretical maximum output power P of the photovoltaic cell array. m =8517.8W, and two local maximum power values ​​P1 =3335.57W and P2 =3359.66W.

[0116] In a partial shadow environment, the MPPT performance of IP&O algorithm, TSO algorithm, SSA algorithm and ISSA-IP&O algorithm are simulated respectively, and the output power characteristic curves are compared. Figure 10 As shown in the figure, in the partial shadow environment, the maximum power output of the TSO algorithm is 4435.26W, and the tracking accuracy is 99.98%; the maximum power output of the SSA algorithm is 4445.35W, and the tracking accuracy is 99.98%; the maximum power output of the ISSA algorithm is 4445.42W, and the tracking accuracy is 99.99%; the maximum power output of the ISSA-IP&O algorithm is 4446.02W, and the tracking accuracy is 99.99%. The experiment shows that the four algorithms can all track the global maximum power point in the partial shadow environment, and the ISSA-IP&O algorithm has the highest optimization accuracy and can find the maximum power point more accurately. The comparison of simulation experimental results is shown in Table (5).

[0117] Table (5) Tracking performance under partial shadow conditions

[0118]

[0119] As can be seen from the data in Table (4), the TSO algorithm tracks the maximum power point and maintains stable output in 0.295s, the SSA algorithm tracks the maximum power point and maintains stable output in 0.313s, the ISSA algorithm tracks the maximum power point and maintains stable output in 0.269s, and the ISSA-IP&O algorithm tracks the maximum power point and maintains stable output in 0.222s. Compared with the tracking time of the TSO, SSA, and ISSA algorithms, the ISSA-IP&O algorithm shortens the tracking time by 24.75%, 29.07%, and 17.47%, respectively.

[0120] Comparison of PU output characteristic curves of photovoltaic cell arrays when light intensity changes under dynamic shadow environment Figure 11 As shown in the figure, in a dynamic shadow environment, when the light intensity of the working environment of the photovoltaic cell array changes, its PU output characteristic curve presents a multi-peak characteristic with three peaks, which are the theoretical maximum output power P of the photovoltaic cell array and the maximum output power P of the photovoltaic cell array. m =5675.41W, and two local maximum power values ​​P3 =3016.53W and P4 =4912.77W.

[0121] In a dynamic shadow environment, the MPPT performance of IP&O algorithm, TSO algorithm, SSA algorithm and ISSA-IP&O algorithm are simulated respectively. Figure 12 As shown in the figure, before the lighting conditions changed, the TSO algorithm tracked the maximum power point at 0.295s, with an output power of 4445.26W; the SSA algorithm tracked the maximum power point at 0.313s, with an output power of 4445.35W; the ISSA algorithm tracked the maximum power point at 0.269s, with an output power of 4445.42W; and the ISSA-IP&O algorithm tracked the maximum power point at 0.222s, with an output power of 4446.02W. After the light intensity changes at 0.4s, the TSO algorithm re-tracks to the maximum power point at 0.701s, with an output power of 5663.16W and a tracking accuracy of 99.78%; the SSA algorithm re-tracks to the maximum power point at 0.701s, with an output power of 5670.06W and a tracking accuracy of 99.91%; the ISSA algorithm re-tracks to the maximum power point at 0.661s, with an output power of 5675.16W and a tracking accuracy of 99.99%; the ISSA-IP&O algorithm re-tracks to the maximum power point at 0.584s, with an output power of 5675.17W and a tracking accuracy of 99.99%. The experimental results show that when the photovoltaic array operates in a dynamic shadow environment, the four algorithms can effectively handle the impact of dynamic shadows, so that the photovoltaic array maintains a stable maximum power output. The comparison of simulation experimental results is shown in Table (6).

[0122] Table (6) Tracking performance under dynamic shadow conditions

[0123]

[0124] From the data in Table (6), we can see that after the light intensity changes (after 0.4s), the TSO algorithm takes 0.301s to re-track to the maximum power point and maintain stable output, the SSA algorithm takes 0.278s to re-track to the maximum power point and maintain stable output, the ISSA algorithm takes 0.261s to re-track to the maximum power point and maintain stable output, and the ISSA-IP&O algorithm takes 0.184s to re-track to the maximum power point and maintain stable output. Compared with the TSO, SSA, and ISSA algorithms, the ISSA-IP&O algorithm shortens the tracking time by 38.87%, 33.81%, and 29.50%, respectively. The ISSA-IP&O algorithm can quickly track to the maximum power point and significantly reduce power oscillations during the tracking process.

[0125] Based on the above experimental results, it can be seen that the use of the ISSA-IP&O algorithm for MPPT control can achieve accurate tracking of the maximum power point, with better tracking accuracy and tracking speed. Compared with the use of TSO, SSA and ISSA algorithms, the tracking accuracy is slightly improved, the tracking time is significantly shortened, and the power oscillation is significantly reduced during the tracking of the maximum power point. The specific results are as follows:

[0126] (1) In a uniform illumination environment, the ISSA-IP&O algorithm shortens the tracking time by 34.90%, 37.62%, and 25.95% compared with the TSO, SSA, and ISSA algorithms, respectively.

[0127] (2) In a partially shadowed environment, the ISSA-IP&O algorithm shortens the tracking time by 24.75%, 29.07%, and 17.47% compared to the TSO, SSA, and ISSA algorithms, respectively.

[0128] (3) In a dynamic shadow environment, after the light intensity changes, the ISSA-IP&O algorithm shortens the tracking time by 38.87%, 33.81%, and 29.50% compared with the TSO, SSA, and ISSA algorithms, respectively.

[0129] (4) The ISSA-IP&O algorithm can quickly and accurately track the maximum power point and maintain stable output in both uniform illumination and partial shadow environments. However, the tracking speed in uniform illumination is much slower than that of the IP&O algorithm. The ISSA-IP&O algorithm solves the problem of the IP&O algorithm falling into local extreme value failure in partial shadow environments, and significantly reduces power oscillations during the maximum power point tracking process, which can effectively improve the efficiency of the photovoltaic cell array.

[0130] like Figure 16The three-dimensional scatter plot of light intensity, temperature, and power at the maximum power point of a solar panel is shown. This plot systematically illustrates the relationship between temperature, light intensity, and output power at the maximum power point (MPPT), revealing the performance response of photovoltaic modules under different environmental conditions. As can be seen from the plot, light intensity has a dominant influence on output power, with power increasing significantly with increasing irradiance, reflecting a typical photocurrent enhancement effect. The effect of temperature, on the other hand, exhibits a certain nonlinearity: within the low to moderate temperature range, power increases slightly, likely due to enhanced carrier mobility; however, in the high temperature range, output power saturates and even slightly decreases with increasing temperature, indicating that the high-temperature-induced drop in open-circuit voltage is the primary factor limiting power increase.

[0131] In order to further verify the ability of various algorithms to track the maximum power point of photovoltaic cells under different temperature environments, this paper sets uniform illumination conditions (G = 1000W / m 2 ), temperature variation simulation experiments were conducted on four MPPT algorithms: TSO, SSA, ISSA, and ISSA-IP&O. The temperature was set from 0°C to 25°C, with each test node set at 5°C. The output power variation trend and tracking performance differences of each algorithm were observed. The power-temperature (PT) curve drawn according to the experimental results is shown in the figure. It can be seen that with the increase in temperature, the output power of all algorithms shows a downward trend, which is consistent with the actual performance of photovoltaic cells. The increase in temperature will cause the open-circuit voltage of photovoltaic cells to decrease, thereby reducing the output power. This is an inherent characteristic determined by the temperature coefficient of silicon material.

[0132] At all temperature points, the ISSA-IP&O algorithm can always track the theoretical maximum power point of the photovoltaic array with extremely high precision, and has good stability and anti-interference capabilities. Its tracking accuracy is always maintained at above 99.99%, and the maximum power output is almost consistent with the theoretical value. At the same time, its tracking time is relatively short, especially in high-temperature environments, and it can complete the maximum power point search more quickly and output stably. The ISSA algorithm is slightly inferior to the ISSA-IP&O in overall tracking accuracy, but it still performs well under medium and low temperature conditions (0℃–15℃), with good stability and tracking time close to that of the ISSA-IP&O. The SSA algorithm has good optimization capabilities at lower temperatures (0℃, 5℃), but as the temperature gradually increases, its tracking accuracy fluctuates to a certain extent, and the output power deviates slightly from the theoretical value.

[0133] As a further refinement, a hybrid algorithm combining an improved sparrow search algorithm and a variable-step perturbation-observation method can be used to track the maximum power point of the PV array when operating in a high-temperature environment; and an improved sparrow search algorithm can be used to track the maximum power point of the PV array under medium- and low-temperature conditions. Specifically, when the environment reaches high temperatures, the hybrid algorithm (ISSA-IP&O) is forced to be used; when the environment reaches medium- and low-temperature temperatures, only the ISSA algorithm is executed, skipping the IP&O phase and not switching to the IP&O algorithm. Adding temperature as a decision condition limits the algorithm execution method under different temperature conditions.

[0134] When the photovoltaic cell array operates in a low temperature environment of 5°C–15°C, an improved sparrow search algorithm is used to track the maximum power point of the photovoltaic array. When the photovoltaic cell array operates in a low temperature environment of 0°C–less than 5°C, a sparrow search algorithm is used to track the maximum power point of the photovoltaic array.

[0135] The performance of the TSO algorithm is significantly inferior to that of the other three optimization algorithms at all temperature points, especially in high-temperature environments (20°C–25°C), where it exhibits large power deviations and obvious oscillations. Its local search and global optimization capabilities are somewhat deficient, making it unsuitable for standalone use in dynamic or high-temperature environments.

[0136]

[0137] In summary, the purpose of the present invention is achieved.

Claims

1. A photovoltaic cell maximum power point tracking method based on an improved sparrow search algorithm, characterized by: The composite algorithm of the improved sparrow search algorithm and the variable step-size perturbation observation method is used as the control algorithm of the MPPT controller. The steps are as follows: at the initial stage of the algorithm operation, the improved sparrow search algorithm is first used. After completing the initialization of the algorithm parameters, it is iterated according to the process of the improved sparrow search algorithm to adjust the duty cycle output of the circuit and calculate the output power of the photovoltaic array under the current duty cycle. The output power is the fitness value of the algorithm and the global optimal position is updated. After the improved sparrow search algorithm has iterated a set number of times, the algorithm has fully explored the entire search space and converged to the vicinity of the global maximum power point, which can be regarded as the maximum power point tracking of the photovoltaic array under uniform illumination conditions. At this time, the improved perturbation observation method is switched to quickly track the maximum power point. The position of the optimal sparrow individual in the improved sparrow search algorithm before the algorithm switching is the initial position for the variable step-size perturbation observation method to search for the optimal value. When the light intensity of the working environment of the photovoltaic array changes, resulting in a significant change in the output power, it is necessary to restart the composite algorithm of the improved sparrow search algorithm and the variable step-size perturbation observation method to track the new maximum power point again. Restart condition formula (1): |PP old |>0.05×P old In formula (1), P represents the current power value, P old Indicates the maximum power value recorded last time; Among them, the specific process of improving the sparrow search algorithm is as follows: S1. Set the initial parameters of the SSA algorithm: population size, search space boundary, safety value and warning value; S2. The chaotic map completes population initialization and divides the sparrow population into discoverers, joiners, and guards according to the set ratio; S3. Calculate the fitness value of each individual according to the corresponding functional relationship; S4. According to formula (2) Formula (3) Formula (4) Complete the location update of the discoverer, joiner and vigilant; In formula (2), t represents the current number of iterations, t max represents the maximum number of iterations; j represents the dimension of the variable to be optimized; X (n,j) Represents the position of the nth sparrow individual in the jth dimension; c is a random number between 0 and 1 and the value of c is not 0. R2 is the warning value of the sparrow population, usually in the range of R2∈[0,1]; S T It is a pre-set safety value, usually S T ∈[0.5,1]; Q is a random number that obeys the normal distribution; A is a matrix whose elements are all 1, the number of rows is 1, and the number of columns is the same as the dimension of the variable; In formula (3), N is the number of sparrow population; X w,t is the position of the individual with the lowest energy in the current sparrow population, X P,t is the position of the individual with the highest energy among the current discoverers; B has the same number of rows and columns as A, where the elements are randomly selected between -1 and 1, and B + =B T (BB T ) -1 ; In formula (4), X b,t is the position with the highest energy in the current sparrow population; b is a random number that follows a normal distribution with a mean of 0 and a variance of 1, which is used to control the step size of the position update of the sentinel; c1 is a random number that represents the direction of the sparrow's movement and controls the step size of the position update, c1∈[-1,1]; f n is the fitness value of the current sparrow individual; f g and f w are the fitness values ​​of the current global best position and the worst position respectively; S5. For individuals that exceed the search boundary, they are processed according to the proposed improved boundary processing strategy to bring them back into the search space. S6. Using the reverse learning strategy according to formula (5) X2 = l + u - X1 to generate a reverse population and calculate the individual fitness value. By comparing the fitness values, the sparrow individual with the higher fitness value is selected to enter the next iteration; S7. Determine whether the algorithm termination condition is met: If the termination condition is met, output the optimal individual and its corresponding fitness value; if the termination condition is not met, return to step S4 and continue running until the termination condition is met.

2. The photovoltaic cell maximum power point tracking method based on the improved sparrow search algorithm according to claim 1, characterized in that: When the photovoltaic array works in a high temperature environment, a composite algorithm of an improved sparrow search algorithm and a variable step size perturbation observation method is used to track the maximum power point of the photovoltaic array; when it works in a medium and low temperature environment, an improved sparrow search algorithm is used to track the maximum power point of the photovoltaic array.

3. The photovoltaic cell maximum power point tracking method based on the improved sparrow search algorithm according to claim 2, characterized in that: When the working environment of the photovoltaic cell array is in the low temperature condition of 5℃–15℃, the improved sparrow search algorithm is used to track the maximum power point of the photovoltaic array. When the working environment of the photovoltaic cell array is in the low temperature condition of 0℃-less than 5℃, the sparrow search algorithm is used to track the maximum power point of the photovoltaic array.

Images