A photovoltaic system maximum power point tracking method

Abstract

The application discloses a photovoltaic system maximum power point tracking method, comprising the following steps: step 1, a base fuzzy disturbance observation method is used to track a nearest point near a power extreme point of a current system output, so as to obtain a voltage value UMPP1 and a current value IMPP1 of the nearest point corresponding to the current power extreme point; step 2, the voltage and current values of the nearest point are used as input parameters, and an improved sparrow algorithm is used to accurately optimize a maximum power point of the photovoltaic system; step 3, whether the current system output power value meets a termination condition is judged; if yes, the current system output power value is not a global maximum power, the search is stopped, and step 4 is executed; otherwise, the current system output power value is a global maximum power point; and step 4, a chaos mapping is used to change a current sparrow position, that is, the current system output power value is changed, and step 3 is executed. The real-time weight is used to enable the sparrow population to track a target function corresponding value more rapidly, so that the global maximum power point can be tracked rapidly under the PSC condition.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of photovoltaic systems, in particular to a photovoltaic system maximum power point tracking method. BACKGROUND

[0002] Photovoltaic solar panels are easy to install, low maintenance cost, long service life, and photovoltaic power generation technology has been widely used in China. In the actual use of photovoltaic components, due to the changeable weather factors, the incident angle of sunlight also changes frequently, resulting in local characteristic curve mutation in photovoltaic power generation process under shadow shielding condition. In the observation of maximum power tracking process, the maximum power point cannot be accurately found. Therefore, the maximum power point tracking under the condition of mutation in photovoltaic power generation is an indispensable point in the research of photovoltaic system. The research on partial shading (PSC) has been carried out worldwide, and has attracted widespread attention from researchers in the field. Therefore, it is very important to study the stable operation of photovoltaic system at the maximum power point and reduce the loss of photovoltaic system devices, which can improve the reliable support for subsequent grid-connected operation of photovoltaic system.

[0003] At present, due to the imperfection of photovoltaic cell materials and its control algorithm, photovoltaic power generation power is greatly affected by external environmental factors, and solar energy is often not fully utilized. And it is difficult to make a breakthrough in a short time. Photovoltaic array is easy to fall into multi-peak situation when running under shadow condition, which is not conducive to system operation. SUMMARY

[0004] The purpose of the present application is to provide a photovoltaic system maximum power point tracking method for the situation that photovoltaic system running under partial shadow condition is easy to fall into local optimum.

[0005] The technical scheme adopted by the present application is:

[0006] A photovoltaic system maximum power point tracking method, comprising the following steps:

[0007] Step 1, based on fuzzy disturbance observation method (FP&O), the nearest point near the power extreme point of the current system output is tracked, and the voltage value UMPP1 and the current value IMPP1 of the nearest point corresponding to the current power extreme point are obtained;

[0008] Step 2, taking the voltage and current values of the nearest point as input parameters, the improved sparrow algorithm is used to accurately optimize the maximum power point of the photovoltaic system;

[0009] Step 3, judge whether the current system output power value meets the termination condition; if yes, the current system output power value is not the global maximum power, stop searching and execute step 4; otherwise, the output power value of the current system is the global maximum power point.

[0010] Step 4, the current sparrow position is changed by using the chaotic mapping, that is, the current system output power value is changed and step 3 is executed.

[0011] Specifically, the sparrow population is subjected to chaotic mapping, and the present application adopts Sinusoidal mapping. Compared with other mappings, Sinusoidal mapping has great improvement in global traversal and optimization efficiency. The characteristics of Sinusoidal mapping make the population in a relatively stable state. If the output value is too different from the target value at the end of the algorithm optimization process, the current sparrow position is changed through Sinusoidal mapping to avoid falling into local optimum. The improved sparrow population can search the maximum voltage value and the maximum current value corresponding to the maximum power point of the system at this time more quickly. If the current sparrow population is at the w1 position, the sparrow population is subjected to formula processing, and the position of the sparrow population Through formula operation, is a random number between 0 and 1, which is used to map the initial spatial position of the finder and the follower. If the previous position of the sparrow is 200 m, the corresponding current optimization power is 200 w. After chaotic mapping, the power is greater than 200 w. If the position of the sparrow falls into a local optimal solution, the advantage of chaotic mapping of the population is embodied, which can quickly get rid of the current position and continue optimization.

[0012] Further, step 1 specifically comprises the following steps:

[0013] Step 1-1, perturbing and observing the photovoltaic array according to the perturbation step length corresponding to the step length coefficient, sampling the output voltage and output current of the photovoltaic system, and calculating the power difference value from the previous sampling;

[0014] Step 1-2, judging whether the power difference value is greater than 0; if yes, the next perturbation direction is positive perturbation, that is, the current input voltage is increased, and step 1-3 is executed; otherwise, the next perturbation direction is reverse perturbation, that is, the current input voltage is reduced and step 1-3 is executed.

[0015] Step 1-3, judging whether the current cycle number is less than the set number of times; if yes, the adjustment step length coefficient is calculated based on the power difference value and step 1-1 is executed; otherwise, the current sampled output voltage and current value are recorded as the voltage and current values of the nearest point near the power extreme point.

[0016] Further, a fuzzy logic factor is introduced in step 1-1 The numerical value is fuzzified to calculate the power difference value ΔP from the previous sampling. The specific calculation formula is as follows:

[0017]

[0018] Wherein, P(t) is the power value at t time, P(t-1) is the power value at t-1 time

[0019] Further, the calculation formula for calculating the adjustment step length coefficient in step 1-3 is:

[0020] λ=α|ΔP|

[0021] Wherein, λ is the adjustment step length coefficient, and α is the step length coefficient.

[0022] Further, in step 2, the improved sparrow algorithm is combined with global chaos and reverse elite learning strategy to improve the convergence speed, and the system output power calculated by the voltage value and the current value corresponding to the maximum power point of the system is taken as the fitness value, and the power maximum value is taken as the objective function for optimization.

[0023] Further, step 2 specifically includes the following steps:

[0024] Step 2-1, the voltage value UMPP and the current value IMPP corresponding to the maximum power point are obtained through the P / V and I / V characteristic curves of the photovoltaic array,

[0025] Step 2-2, the voltage value UMPP1 and the current value IMPP1 tracked by the fuzzy disturbance observation method (FP&O) are multiplied to obtain the power value of the nearest point;

[0026] Step 2-3, based on the power value of the nearest point, the current optimization power value is searched by using the improved sparrow algorithm;

[0027] Step 2-4, the current optimization power value is compared with the system maximum power value to determine whether the population fitness falls into the local optimal solution value corresponding to the population; if yes, the previous optimization power value is abandoned and the optimization is returned to step 2-3; otherwise, step 2-5 is executed;

[0028] Step 2-5, the current optimization power value is compared with the set elite sparrow population to preliminarily determine whether the current optimization power value is close to the allowable range of the maximum power point; if yes, step 2-7 is executed; otherwise, the fitness value is updated and step 2-6 is executed;

[0029] Step 2-6, the population iteration number is updated, the position of the sparrow population is updated so as to approach the maximum power direction, the position is updated by using the position update formula, and step 2-4 is executed

[0030] The position update formula of the predator is:

[0031]

[0032] Wherein, t is the current running iteration number; j is the dimension of the optimization target; imax is the maximum number of iterations; a and Q are random numbers between [0, 1] that conform to normal distribution; is the sparrow position of the tth optimization, and L is a 1-row and j-column matrix with all 1s.

[0033] The position updating formula of the companion is:

[0034]

[0035] wherein, is the position corresponding to the lowest fitness of this iteration; A + = A T (A·A i ) -1 A is a 1-row and S-column matrix, and the elements in the matrix are assigned values of 1 or -1; when the number of companions is less than the population number n / 2, it indicates that the current companion is in a state of hunger, and the search range is foraging. The companion supervises the predator during the tracking process, and when a more superior foraging range is found, it will compete with the predator, update the position if the competition is successful, and continue to follow if the competition fails.

[0036] The position updating formula of the anti-predator is:

[0037]

[0038] wherein, β is a step size coefficient satisfying a number between [0, 1] that conforms to normal distribution, used to control the distance of the next search range, K takes values [-1, 1]; ε is a very small value; f i is the fitness corresponding to the ith population, f g corresponds to the current optimal fitness value.

[0039] Step 2-7, perform adaptive T-distribution mutation operation on the current optimization power value, and judge whether the system is in a local shielding condition; if yes, return to step 2-6; otherwise, output the duty cycle corresponding to the current optimization power value.

[0040] Step 2-8, control the boost circuit for voltage and current value control through the duty cycle, so as to adjust the output power of the photovoltaic array and make the photovoltaic array stably operate at the maximum power point.

[0041] Further, step 3 specifically includes the following steps:

[0042] Step 3-1, respectively acquire the output power value P t of the system at time t and the output power value P t+1 at time t+1,

[0043] Step 3-2, calculate and judge If yes, the output power value of the current system is the global maximum power point; otherwise, the output power value of the current system is not the global maximum power, the search is stopped and step 1 is performed.

[0044] The present application adopts the above technical solution, firstly uses the fuzzy logic observation method to track the maximum power point of the system, and then uses the improved war strategy algorithm to accurately find the current maximum power point of the system. The improved war strategy algorithm is not easy to fall into local optimum when tracking power under shadow condition and timely responds to the I / V change of the system. The present application combines the fuzzy logic disturbance and the improved sparrow algorithm. When the system runs, the logic disturbance observation method is used to quickly track the maximum power point of the system. The chaos strategy is used to effectively avoid falling into local optimal solution in the tracking process. The elite reverse learning strategy is added to the traditional sparrow algorithm, so that the sparrow population will not deviate far from the foraging range in the foraging process. The present application uses real-time weight to make the sparrow population quickly track the target function corresponding value, so that the global maximum power point can be quickly tracked under PSC condition. BRIEF DESCRIPTION OF DRAWINGS

[0045] The present application will be further described in detail below in combination with the drawings and specific embodiments.

[0046] Figure 1 The figure is a flowchart of the maximum power point tracking method of the photovoltaic system of the present application.

[0047] Figure 2 The figure is a P-V characteristic curve diagram.

[0048] Figure 3 The figure is an output power comparison diagram under the condition that the irradiance is constant.

[0049] Figure 4 The figure is an output power comparison diagram under local shading.

[0050] Figure 5 The figure is a P-V characteristic curve diagram under irradiance change.

[0051] Figure 6 The figure is an I-V characteristic curve diagram under irradiance change.

[0052] Figure 7 The figure is an output power comparison diagram under variable temperature and irradiance.

[0053] Figure 8 The figure is an output power comparison diagram under variable temperature and irradiance. EMBODIMENT

[0054] In order to make the purpose, technical scheme and advantages of the embodiments of the present application more clear, the technical scheme of the embodiments of the present application will be described clearly and completely in combination with the drawings of the embodiments of the present application.

[0055] As Figures 1 to 8 The application discloses a photovoltaic system maximum power point tracking method, which comprises the following steps:

[0056] Step 1, the nearest point near the power extreme point of the current system output is tracked based on a fuzzy perturbation and observation method (FP&O), so as to obtain a voltage value UMPP1 and a current value IMPP1 of the nearest point corresponding to the current power extreme point;

[0057] Step 2, the voltage and current values of the nearest point are used as input parameters, and an improved sparrow algorithm is used to accurately optimize the maximum power point of the photovoltaic system;

[0058] Step 3, whether the current system output power value meets a termination condition is judged; if yes, the current system output power value is not a global maximum power, the search is stopped and step 4 is executed; otherwise, the output power value of the current system is a global maximum power point.

[0059] Step 4, the current sparrow position is changed by using a chaotic mapping, that is, the current system output power value is changed and step 3 is executed.

[0060] Specifically, the sparrow population is subjected to chaotic mapping, and the application adopts a Sinusoidal mapping. Compared with other mappings, the Sinusoidal mapping has great improvement in global traversal and optimization efficiency. The characteristics of the Sinusoidal mapping make the population in a relatively stable state. If the output value and the target value are too different at the end of the algorithm optimization process, the current sparrow position is changed by the Sinusoidal mapping, so as to avoid falling into a local optimum. The expression of the Sinusoidal mapping is:

[0061]

[0062] S (i+1) =w (i+1) ×S (i) (2-3)

[0063] In the formula, w (i) is the current sparrow position, S (i+1) is the population position after mapping, w (1) =0.8,

[0064] The improved sparrow population can more quickly search the maximum voltage value and the maximum current value corresponding to the maximum power point of the system at this time. If the current sparrow population is in a w1 position, the sparrow population is subjected to formula processing, and the sparrow population is in a w1× Through formula operation, A random number between 0 and 1 is used to map the initial spatial positions of the discoverer and the follower. If the sparrow's previous position was 200m, the current optimization power was 200w. After chaotic mapping, the power is greater than 200w. If the sparrow's position is trapped in a local optimum, the benefit of chaotic mapping on the population becomes apparent, as it can quickly get rid of its current position and continue to optimize.

[0065] Furthermore, step 1 specifically includes the following steps:

[0066] Step 1-1: Perform perturbation observation on the photovoltaic array according to the perturbation step size corresponding to the step size coefficient, sample the output voltage and output current of the photovoltaic system, and calculate the power difference with the previous sample.

[0067] Step 1-2: Determine if the power difference is greater than 0; if so, the next disturbance direction is a positive disturbance, i.e., increase the current input voltage, and execute step 1-3; otherwise, the next disturbance direction is a negative disturbance, i.e., decrease the current input voltage and execute step 1-3.

[0068] Steps 1-3: Determine if the current number of iterations is less than the set number of iterations; if so, calculate and adjust the step size coefficient based on the power difference and execute step 1-1; otherwise, record the currently sampled output voltage and current values ​​as the voltage and current values ​​of the nearest point near the power extreme point.

[0069] Furthermore, fuzzy logic factors are introduced in step 1-1. The numerical values ​​are fuzzified to calculate the power difference ΔP with the previous sample. The specific calculation formula is as follows:

[0070]

[0071] Where P(t) is the power value at time t, and P(t-1) is the power value at time t-1.

[0072] Furthermore, the calculation formula for the adjustment step size coefficient based on the power difference in steps 1-3 is as follows:

[0073] λ=α|ΔP|

[0074] Where λ is the adjustment step size coefficient and α is the step size coefficient.

[0075] Furthermore, in step 2, the improved sparrow algorithm combines global chaos and reverse elite learning strategies to improve the convergence speed, and uses the system output power calculated by the voltage and current values ​​corresponding to the system's maximum power point as the fitness value, and uses the maximum power value as the objective function for optimization.

[0076] Furthermore, step 2 specifically includes the following steps:

[0077] Step 2-1: Obtain the voltage value UMPP and current value IMPP corresponding to the maximum power point using the P / V and I / V characteristic curves of the photovoltaic array.

[0078] Step 2-2: Multiply the voltage value UMPP1 and the current value IMPP1 tracked by the fuzzy perturbation observation method (FP&O) to obtain the power value of the nearest point;

[0079] Steps 2-3: Based on the power value of the nearest point, use the improved sparrow algorithm to search for the current optimal power value;

[0080] Step 2-4: Compare the current optimization power value with the system's maximum power value to determine whether the population fitness is trapped in the corresponding local optimum. If so, abandon the previous optimization power value and return to step 2-3 to optimize again; otherwise, proceed to step 2-5.

[0081] Step 2-5: Compare the current optimization power value with the set elite sparrow population to preliminarily determine whether the current optimization power value is close to the maximum power point within the allowable range; if so, proceed to step 2-7; otherwise, update the fitness value and proceed to step 2-6.

[0082] Steps 2-6: Population iteration count update. Update the position of the sparrow population to move towards the direction of maximum power. Update the position using the position update formula, and then execute steps 2-4.

[0083] The predator's position update formula is:

[0084]

[0085] Where t is the current iteration number; j is the dimension of the optimization objective; i max It is the maximum number of iterations; α and Q are random numbers between [0,1] that follow a normal distribution. Let L be the position of the sparrow in the t-th optimization, and L be a 1-row, j-column matrix consisting entirely of 1s.

[0086] The position update formula for the companion is:

[0087]

[0088] in, This is the position corresponding to the lowest fitness in this iteration; A + =A T (A·A T ) -1Let A be a 1-row, S-column matrix, with each element having a value of 1 or -1. When the number of companions is less than the population size n / 2, it indicates that the current companion is hungry and searches for food. During the tracking process, the companion monitors the predator. When it discovers a more favorable foraging area, it will compete with the predator. If the competition is successful, its position is updated. If the competition fails, it continues to follow.

[0089] The formula for updating the position of the anti-predator is:

[0090]

[0091] Where β is a step size coefficient that follows a normal distribution and is between [0, 1], used to control the size of the next search range; K takes values ​​between [-1, 1]; ε is a local minimum; f i f represents the fitness of the i-th population. g This corresponds to the current optimal fitness value.

[0092] Step 2-7: Perform an adaptive T-distribution mutation operation on the current optimization power value to determine whether the system is under local shading conditions; if so, return to step 2-6; otherwise, output the duty cycle corresponding to the current optimization power value.

[0093] Steps 2-8 involve controlling the voltage and current values ​​of the boost circuit by controlling the duty cycle, thereby adjusting the output power of the photovoltaic array and ensuring that the photovoltaic array operates stably at its maximum power point.

[0094] Furthermore, step 3 specifically includes the following steps:

[0095] Step 3-1: Obtain the output power value P of the system at time t. t The output power value P at time t+1 t+1 ,

[0096] Step 3-2, calculate and judge Is it true? If yes, the current system output power value is the global maximum power point; otherwise, if the current system output power value is not the global maximum power, stop the search and execute step 1.

[0097] The specific principles of this invention will be explained in detail below:

[0098] (1) The maximum power point is pre-tracked by the fuzzy perturbation observation method and the current output power of the photovoltaic array is obtained by setting the relevant parameters of the improved sparrow algorithm.

[0099] Specifically, based on the fuzzy perturbation and observation (FP&O) method, the system tracks the area near the current system output power extremum. Under constant external conditions, the PV battery system has only one maximum power point, which is also the global maximum power point. This theoretical basis provides the implementation conditions for fuzzy perturbation and observation P&O. In the early stages of algorithm operation, perturbation is introduced into the input signal, and the system detects the current direction and parameters. After the system outputs the current state, the perturbation input to the system is adjusted according to preset parameters to control the algorithm to approach the preset value. This is the essence of perturbation and observation method, which applies step observation to the MPPT direction.

[0100] If the system voltage UMPP does not intersect with the current power point, and the intersection point is to the left of the characteristic curve, the algorithm will increase the voltage at the intersection point. Conversely, if the intersection point is to the right of the characteristic curve, the algorithm will decrease the voltage at the intersection point. This process is repeated until the system is near the maximum power point.

[0101] like Figure 2 As shown, when the algorithm operates to the left of the maximum power point, the current voltage value is adjusted to U + ΔU; conversely, when the system operates to the right of the maximum power point, the current voltage value is adjusted to U - ΔU. The voltage difference before and after the output power is calculated to determine the direction of the next disturbance. A positive ratio indicates the disturbance direction is the same as the current one, while a negative ratio indicates the opposite. It should be noted that the disturbance step size of the fuzzy disturbance observation method is 0.1, and a fuzzy logic factor is introduced. To blur the numerical values. The value is between 0.1 and 1. By adding a fuzzy factor, it can approach the current maximum power point more quickly. Record the output voltage and current values ​​after the perturbation observation method, UMPP = UMPP1, IMPP = IMPP1, and send the voltage and current values ​​searched by the fuzzy perturbation observation method into the improved sparrow algorithm to further accurately optimize the corresponding maximum power point of the system.

[0102] (2) A relatively stable group of sparrows is established within the optimization range using chaotic mapping. Specifically, to divide the discoverer and follower groups in the population according to a certain ratio, each sparrow randomly selects its own identity, such as predator or discoverer. Specifically, to accurately search for the system's maximum voltage point UMPP and maximum current point IMPP, thereby improving the search for the maximum power point, a chaotic mapping is used on the sparrow population. This invention uses the Sinusoidal mapping, which significantly improves global traversal and optimization efficiency compared to other mappings. The characteristics of the Sinusoidal mapping keep the population in a relatively stable state. If the current sparrow population is at position w1, the sparrow population is processed using an arithmetic formula to determine its position. Through formula manipulation is a random number between 0 and 1, used to map the initial spatial positions of discoverers and followers. If the previous position of the sparrow was 200m and the corresponding current optimization power was 200w, after chaotic mapping, the power is greater than 200w. If the position of the sparrow falls into a local optimal solution, the advantage of chaotic mapping for the population is manifested, enabling it to quickly break away from the current position and continue the optimization process.

[0103] (3) Calculate the output power of the system by the voltage value and current value corresponding to the maximum power point of the system, determine its fitness value, and take the maximum value of the power as the objective function of the algorithm. From the P / V and I / V characteristic curves of the photovoltaic array, the voltage value UMPP and current value IMPP corresponding to the maximum power point can be known. Multiply the voltage value UMPP1 and current value IMPP1 tracked by the fuzzy perturbation observation method (FP&O) to obtain the values around the maximum power point. Further, compare the current power value searched by the improved sparrow algorithm with the value of the maximum power point of the system.

[0104] (4) Compare the fitness values, comparing the individual and global extreme values. Specifically, from step (2), the power values corresponding to the current optimization of the sparrow population are determined, and the optimized power value is its fitness value. Compare the current corresponding power value with the maximum power point of the system, compare the fitness values of all current sparrow individuals, and retain the optimal fitness value and the best position. The optimal value corresponding to the objective function is set as fmax, fmax is the value of the maximum power point in the pv characteristic curve, f is the value of the objective function optimized by the current sparrow. If f > fbest, then Xbest = X corresponding to which class of the current sparrow population is optimized, fbest = f. If f < fbest, then retain f.

[0105] Secondly, an elite sparrow population is added to the traditional sparrow algorithm. Increasing the elite reverse population can improve the search accuracy and search range of the population, and the algorithm is not easily trapped in local optima. The formula for the elite sparrow population is (2 - 4):

[0106]

[0107] In the formula: is the solution of the elite population, is the value of the current position of the soldier, u b 、l b are the initially set upper and lower limit values, K ∈ (0,1), and is a randomly distributed number.

[0108] First, the optimization power of the ordinary sparrow population is compared with the optimized power of the elite sparrow population. If the optimization power of the ordinary sparrow population is less than that of the elite sparrow population, the output is the power optimized by the elite sparrow population, and vice versa. Adding the elite sparrow population can initially determine whether the current system power value is close to the maximum power point. Furthermore, adding the elite reverse population can improve the search accuracy and search range of the population, and the algorithm is less likely to get trapped in local optima. If the optimized system power is less than the current system maximum power point value, the fitness value is updated. The population iteration count is updated, and the position of the sparrow population is updated. The sparrow population position update formula is (1-1), (1-2), (1-3). After updating the position, it is moved closer to the direction of maximum power, and the position is updated using the position update formula. The above process is repeated, and the algorithm continues to optimize.

[0109] The predator's position update formula is:

[0110]

[0111] Where t is the current iteration number; j is the dimension of the optimization objective; i max It is the maximum number of iterations; α and Q are random numbers between [0,1] that follow a normal distribution. Let L be the position of the sparrow in the t-th optimization, and L be a 1-row, j-column matrix consisting entirely of 1s.

[0112] The position update formula for the companion is:

[0113]

[0114] in, This is the position corresponding to the lowest fitness in this iteration; A + =A T (A·A T ) -1 Let A be a 1-row, S-column matrix, with each element having a value of 1 or -1. When the number of companions is less than the population size n / 2, it indicates that the current companion is hungry and searches for food. During the tracking process, the companion monitors the predator. When it discovers a more favorable foraging area, it will compete with the predator. If the competition is successful, its position is updated. If the competition fails, it continues to follow.

[0115] The formula for updating the position of the anti-predator is:

[0116]

[0117] Where β is a step size coefficient that follows a normal distribution and is between [0, 1], used to control the size of the next search range; K takes values ​​between [-1, 1]; ε is a local minimum; f i f represents the fitness of the i-th population.g This corresponds to the current optimal fitness value.

[0118] (5) Determine if the conditions are met. If the conditions are met, output the duty cycle corresponding to the current power. If not, continue to step (4) for iteration. Specifically, if the system power optimized by the algorithm is close to the maximum power point, the current maximum power point can be initially determined as the system's maximum power point. Further, perform an adaptive T-distribution mutation operation on the current system power. Because performing a certain probability T-distribution mutation on the power corresponding to the current population position can effectively determine whether the system is under local shading conditions. If the system is under shading conditions, this operation will cause the previous probability value to increase rapidly, which does not match the maximum power point of the objective function. Therefore, it is necessary to return to step (4) to optimize to the corresponding maximum power point of the system. If the algorithm does not increase rapidly after the T-distribution mutation and the value is stable, output the duty cycle corresponding to the current output power. The duty cycle is used to control the voltage and current values ​​of the boost circuit to adjust the output power of the photovoltaic array and make the photovoltaic array operate stably at the maximum power point.

[0119] Specifically, a T-distribution mutation with a certain probability is applied to the power corresponding to the current position of the population to improve the convergence speed of the algorithm. The update formula is as follows:

[0120]

[0121] In the formula: The formula after position update. Let t be the position of the (t-1)th iteration. iter (t) represents the number of iterations that satisfy the T-distribution.

[0122] (6) To more accurately track the global maximum power point, a termination condition is set so that the algorithm restarts when the parameters change. Furthermore, P is set... t+1 P is the current system output power value at time t+1. t It is the current system output power value at time t. If P t+1 -P t power value divided by P t If the value is less than 0.1, the algorithm is considered to be trapped in a local optimum. The global maximum power point sought by the system is not the current corresponding value. The algorithm stops searching and resets to step 1 to restart.

[0123] Therefore, when tracking the maximum power point of a photovoltaic system, the Fuzzy Logic Perturbation-Based Composite Improved Sparrow Algorithm (FP&O-MSSA) provided in this invention utilizes fuzzy logic perturbation to quickly track the approximate maximum power point of the system beforehand. The system parameters are then passed to multiple improved sparrow algorithms for simulation. A combined global chaos and reverse elite learning strategy improves convergence speed, and adaptive weight coordination avoids the system getting trapped in multiple peak values. During system operation, it can respond promptly to changes in I / V. It also improves upon the sparrow algorithm's tendency to encounter local optima under shaded conditions, thus enhancing tracking capability. Comparisons demonstrate the improved convergence speed and tracking accuracy of the improved composite algorithm in tracking the maximum power point.

[0124] To demonstrate the effectiveness of the algorithm of this invention, a simulation model was built in MATLAB / Simulink, consisting of a solar panel, a Boost module, an MPPT module, and a load component. The algorithm of this invention, ISSA, PSO, P&O, and WSO algorithms were compared under different external conditions. To ensure fairness in the experiment, the population size was uniformly set to 10, and the duty cycles were uniformly set to 0.1, 0.3, 0.5, 0.7, and 0.9, with a maximum of 20 iterations. Specifically, the PSO algorithm parameters were set as W = 0.9, C1 = C2 = 2, and the P&O step size was 0.1. The ISSA algorithm parameters were set as follows: population size 10, ST = 0.8, the number of alerters and discoverers was 0.2 of the population, and the algorithm of this invention had a weight of 0.9, with the weight changing exponentially. The lower bound of the algorithm was -100 / W, and the upper bound was 700 / W. The average value of each algorithm was compared after running 20 times.

[0125] The irradiance of the four solar panels was set to [1000, 800, 800, 500] respectively, and the temperature was uniformly set to 25℃. Under these conditions, the tracking efficiency of each algorithm was observed.

[0126] Depend on Figure 3 It is known that the current system's maximum power is 613.25W. After the system has been running for a certain period of time, all algorithms can find the system's maximum power point, but there are still significant differences among the algorithms. The PSO algorithm exhibits oscillating optimization during the tracking process. The P&O algorithm lags in its optimization speed during the tracking of the maximum power point, and the power coefficient varies considerably in a short period of time, which is detrimental to the stable operation of the system. The WSO and ISSA algorithms only show small fluctuations in power during the optimization process and can operate smoothly. Through comparison, it is found that the algorithm of this invention has the most stable optimization process and the fastest convergence speed under constant irradiance.

[0127] Initially, the irradiance of the four solar panels was set to [1000, 800, 800, 500]. Simulations showed that the irradiance jumped to [900, 700, 700, 500] at 0.7 seconds and to [500, 500, 500, 200] at 1.3 seconds. These irradiance jumps were used to illustrate the accuracy of the algorithm in tracking the maximum power point under partial shading conditions.

[0128] Depend on Figure 4 As can be seen, when the irradiance changes abruptly, the algorithm of this invention, compared with several other algorithms, shows a clear difference in tracking the maximum power after the change. When the system irradiance becomes [900, 700, 700, 500], the current maximum power point of the system is 539.39W. All the compared algorithms get stuck in local optima when tracking the maximum power point when the irradiance changes. The algorithm of this invention can escape local optima and accurately track the maximum power point, verifying the effectiveness of the algorithm of this invention.

[0129] In actual operation, the temperature of a photovoltaic system changes with external conditions, affecting the irradiance. To comprehensively test the feasibility of the algorithm, this invention tests the effect of FP&O-ISSA tracking maximum power under varying temperature and irradiance conditions. The initial system settings are: temperature 25℃, irradiance [1000 800 800 500]. After 0.6 seconds, the temperature changes to 50℃, and the irradiance is [900 700 700 500]. After 1.3 seconds, the temperature changes to 60℃, and the irradiance is [1000 1000 800 800]. Figure 5 and 6 The figures show the PV / IV curves of the system under varying temperature and irradiation conditions.

[0130] Table 1 Algorithm Comparison Table

[0131]

[0132]

[0133] This invention employs the above technical solutions. First, it uses fuzzy logic observation to pre-track the vicinity of the system's maximum power point. Then, it uses an improved war strategy algorithm to accurately find the current maximum power point of the system. The improved war strategy algorithm is less prone to getting trapped in local optima when tracking power under shadow conditions and responds promptly to changes in the system's I / V. This invention combines fuzzy logic perturbation with an improved sparrow algorithm. During system operation, it first uses a logic perturbation observation method to quickly track the vicinity of the system's maximum power point. A chaotic strategy is used to effectively avoid getting trapped in local optima during the tracking process. An elite back-learning strategy is added to the traditional sparrow algorithm, ensuring that the sparrow population does not deviate too far from its foraging range during foraging. This invention utilizes real-time weights to enable the sparrow population to track the corresponding value of the objective function more quickly, allowing for rapid tracking of the global maximum power point under PSC conditions.

[0134] Obviously, the described embodiments are only a part of the embodiments of this application, not all of them. Without conflict, the embodiments and features in the embodiments of this application can be combined with each other. The components of the embodiments of this application described and illustrated herein can generally be arranged and designed in various different configurations. Therefore, the detailed description of the embodiments of this application is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

Claims

1. A method of maximum power point tracking for a photovoltaic system, the method comprising: It comprises the following steps: Step 1, based on fuzzy disturbance observation method to track the current system output power extreme point near the nearest point, get the current power extreme point corresponding to the nearest point voltage value UMPP1 and current value IMPP1; Step 1 specifically includes the following steps: Step 1-1, according to the perturbation step length corresponding to the step length coefficient, the perturbation observation is carried out on the photovoltaic array, the output voltage and output current of the photovoltaic system are sampled, and the power difference value is calculated compared with the previous sampling; Step 1-2, judge whether the power difference value is greater than 0; if yes, the next perturbation direction is positive perturbation, that is, to increase the current input voltage, and execute step 1-3; otherwise, the next perturbation direction is reverse perturbation, that is, to reduce the current input voltage and execute step 1-3; Step 1-3, judge whether the current cycle number is less than the set number; if yes, calculate the adjustment step length coefficient based on the power difference value and execute step 1-1; otherwise, record the current sampled output voltage and current value as the voltage and current value of the nearest point near the power extreme point; Step 2, taking the voltage and current value of the nearest point as the input parameter, the improved sparrow algorithm is used to accurately optimize the maximum power point of the photovoltaic system; the improved sparrow algorithm is combined with global chaos and reverse elite learning strategy to improve the convergence speed, and the system output power calculated by the voltage value and current value corresponding to the system maximum power point is taken as the fitness value, and the power maximum value is taken as the objective function to optimize; specifically includes the following steps: Step 2-1, the voltage value UMPP and current value IMPP corresponding to the maximum power point are obtained through the P / V, I / V characteristic curve of the photovoltaic array, Step 2-2, the voltage value UMPP1 and current value IMPP1 tracked by the fuzzy disturbance observation method are multiplied to obtain the power value of the nearest point; Step 2-3, based on the power value of the nearest point, the improved sparrow algorithm is used to search for the current optimized power value; Step 2-4, compare the current optimized power value with the system maximum power value, judge whether the population fitness falls into the corresponding population local optimal solution value; if yes, give up the previous optimized power value and return to step 2-3 to optimize again; otherwise, execute step 2-5; Step 2-5, compare the current optimized power value with the set elite sparrow population to preliminarily judge whether the current optimized power value is close to the allowable range of the maximum power point; if yes, execute step 2-7; otherwise, update the fitness value and execute step 2-6; Step 2-6, update the iteration number of the population, update the position of the sparrow population so as to approach the maximum power direction, update the position by using the position update formula, and execute step 2-4; Step 2-7, perform adaptive T distribution mutation operation on the current optimized power value, judge whether the system is under local shading condition; if yes, return to step 2-6; otherwise, output the duty cycle corresponding to the current optimized power value; Step 2-8, control the boost circuit through the duty cycle to control the voltage and current value, so as to adjust the output power of the photovoltaic array and make the photovoltaic array stably operate at the maximum power point; Step 3, judging whether the current system output power value meets the termination condition; if yes, the current system output power value is not the global maximum power, stopping searching and executing step 4; otherwise, the current system output power value is the global maximum power point; Step 4, changing the current sparrow position by using the chaotic mapping, i.e. changing the current system output power value and executing step 3.

2. The photovoltaic system maximum power point tracking method of claim 1, wherein: Introducing the fuzzy logic factor in step 1-1 The values are fuzzified to calculate the power difference ΔP from the previous sample, with the following formula: Wherein, P(t) is the power value at t moment, P(t-1) is the power value at t-1 moment.

3. The photovoltaic system maximum power point tracking method of claim 1, wherein: The calculation formula for calculating the adjustment step length coefficient in steps 1-3 based on the power difference value is: λ=α|ΔP| Wherein, λ is the adjustment step length coefficient, and α is the step length coefficient.

4. The photovoltaic system maximum power point tracking method of claim 1, wherein: When updating the position in step 2-6, The position updating formula of the predator is: Wherein, t is the current running iteration number; j is the dimension of the optimization target; i max is the maximum number of iterations; a and Q are random numbers between [0, 1] that conform to normal distribution; is the sparrow position of the tth optimization, and L is a 1-row j-column matrix all of which are 1. The position updating formula of the companion is: wherein, is the position corresponding to the lowest fitness of the current iteration; A + = A T (A·A T ) -1 , A is a matrix of 1 row and S columns, the elements in the matrix are assigned values of 1 or -1; n is the population size; The position updating formula of the counter-predator is: Wherein, β is a step coefficient satisfying a normal distribution of a number between [0, 1], used to control the distance of the next search range, K takes [-1, 1]; ε is a minimum value; f i is the fitness value corresponding to the ith population, f g is the current optimal fitness value.

5. The photovoltaic system maximum power point tracking method of claim 1, wherein: Step 3 specifically includes the following steps: Step 3-1, obtain the output power value P of the system at time t t and the output power value P at time t+1 t+1 , Step 3-2, calculate and determine whether it is true; if yes, the output power value of the current system is the global maximum power point; otherwise, the output power value of the current system is not the global maximum power stop searching and perform step 1.

6. The photovoltaic system maximum power point tracking method of claim 1, wherein: In step 4, the chaotic mapping of the sparrow population adopts Sinusoidal mapping.

Images