Photovoltaic maximum power point tracking method based on adaptive vulture search algorithm

Abstract

The photovoltaic maximum power point tracking method based on the adaptive vulture search algorithm includes the following steps: Step 1: Based on the BES algorithm, a Gaussian mixture adaptive walk strategy, a progressive dive adaptive switching strategy, and a vulture flock size adjustment mechanism are introduced to construct the ABES algorithm; Step 2: The ABES algorithm constructed in Step 1 is used for photovoltaic maximum power point tracking; Step 3: Based on Step 2, a photovoltaic power generation system simulation model is built to verify the tracking performance of the ABES algorithm constructed in Step 1 under different scenarios. This invention introduces a Gaussian mixture adaptive walk strategy, a progressive dive adaptive switching strategy, and a vulture flock size adjustment mechanism into the BES algorithm, and applies them to photovoltaic maximum power point tracking. This method can successfully track the global maximum power point under local shading conditions, and the tracking speed is faster and the accuracy is higher.

Description

Technical Field

[0001] This invention relates to the field of maximum power point tracking (MPPT) technology for photovoltaic systems, and specifically to a photovoltaic MPPT method based on an adaptive vulture search algorithm. Background Technology

[0002] The ever-increasing demand for energy and the gradual depletion of fossil fuels have made renewable energy a focus of attention. Solar energy, as a significant source of renewable energy, is characterized by its sustainability and lack of pollution. Photovoltaic (PV) power generation is one of the main methods for large-scale development and utilization of solar energy. However, due to the non-linear nature of the PV characteristic curve of a PV system, it is difficult to ensure maximum output power under varying weather conditions. Therefore, maximum power point tracking (MPPT) technology can be used to track the maximum power point of the PV system, thereby increasing its power generation.

[0003] Existing MPPT technologies are mainly divided into traditional MPPT algorithms and heuristic intelligent algorithms. Traditional MPPT algorithms primarily include perturbation and observation (P&O), incremental conductivity (INC), and hill climbing (HC) methods. These algorithms perform well when irradiance is uniform, i.e., the PV characteristic curve has a single peak. However, photovoltaic arrays are susceptible to partial shading conditions (PSC) caused by clouds and buildings, resulting in a multi-peak PV characteristic curve. In such cases, traditional MPPT algorithms struggle to track the global maximum power point.

[0004] To address the issue of traditional MPPT algorithms failing under partial shading conditions, many researchers have discovered that MPPT techniques based on heuristic intelligent algorithms can more accurately track the global maximum power point when dealing with multi-peak phenomena caused by partial shading. Examples include MPPT techniques for photovoltaic systems using particle swarm optimization (PSO), gray wolf optimization (GWO), differential evolution (DE), and the squirrel search algorithm (SSA).

[0005] Existing research mainly introduces novel heuristic intelligent algorithms to improve MPPT tracking performance. The Bald Eagle Search (BES) algorithm, as a novel heuristic intelligent algorithm, has strong global search capabilities and can effectively solve various complex numerical optimization problems. However, in photovoltaic maximum power point tracking applications, this algorithm has problems such as slow convergence speed, easy getting trapped in local optima, and incoordination between global search and local development. Summary of the Invention

[0006] To address the issue of traditional MPPT algorithms failing under partial shading conditions, this invention proposes a photovoltaic maximum power point tracking (MPPT) method based on the ABES algorithm. This method introduces a Gaussian mixture adaptive walk strategy, a progressive dive adaptive switching strategy, and a bald eagle flock size adjustment mechanism into the BES algorithm. When applied to photovoltaic MPPT, it can successfully track the global maximum power point under partial shading conditions, with faster tracking speed and higher accuracy.

[0007] The technical solution adopted in this invention is as follows:

[0008] The photovoltaic maximum power point tracking method based on the adaptive vulture search algorithm includes the following steps:

[0009] Step 1: Based on the BES algorithm, introduce the Gaussian mixture adaptive walk strategy, the progressive dive adaptive switching strategy, and the eagle flock size adjustment mechanism to construct the ABES algorithm;

[0010] Step 2: Apply the ABES algorithm constructed in Step 1 to photovoltaic maximum power point tracking;

[0011] Step 3: Based on Step 2, build a simulation model of the photovoltaic power generation system to verify the tracking performance of the ABES algorithm constructed in Step 1 under different scenarios.

[0012] In step 1,

[0013] A Gaussian mixture adaptive walk strategy is introduced in the selection phase of the BES algorithm. The iterative formula for the selection phase of the ABES algorithm is as follows:

[0014]

[0015]

[0016] In the formula: This represents the position of the i-th vulture at the (t+1)-th iteration; This indicates the optimal position of the vulture flock at the t-th iteration; This indicates the center position of the vulture flock at the t-th iteration; N(μ,σ) represents the position of the i-th bald eagle in the t-th iteration;2 ) indicates that it follows a sequence with expected value of μ and variance of σ. 2 Gaussian distributed random variable; t max This indicates the maximum number of iterations.

[0017] ε1 represents the variance. The weights of the Gaussian distribution; ε² represents the variance. The weights of the Gaussian distribution; t represents the number of iterations; σ1 represents the standard deviation of the first Gaussian distribution; σ2 represents the standard deviation of the second Gaussian distribution.

[0018] A progressive dive adaptive switching strategy is introduced in the dive phase of the BES algorithm. The iterative formula for the dive phase of the ABES algorithm is as follows:

[0019]

[0020] γ=r2-p s ;

[0021] In the formula, H(·) represents the Herveside step function; Sgn(·) represents the sign function; x1(i) and y1(i) represent the position of the vulture in polar coordinates; c1 and c2 represent the intensity of the vulture's movement towards the optimal and central positions; r1 and r2 both represent random numbers between (0,1); γ represents the fish's escape state; p s This indicates the probability that the fish successfully escaped.

[0022] p s This represents the probability that the fish successfully escapes. Its value decreases non-linearly from 0.75 to 0 as the number of iterations increases, using an adaptive sinusoidal decay method. The specific formula is as follows:

[0023]

[0024] When γ < 0, fish have a chance to escape successfully, while the bald eagle flock adopts a gradual dive motion;

[0025] When γ>0, the fish have no chance of escaping, and the eagle flock will dive.

[0026] Considering that the intensity of a vulture's movement weakens over time, this invention proposes a nonlinear adaptive attenuation factor that takes into account the dynamic nature of the vulture's movement intensity. The specific formula is as follows:

[0027]

[0028] In the formula, c min and c max These are the minimum and maximum values ​​of parameters c1 and c2 within the range [1,2], respectively; β is the contraction coefficient.

[0029] A dynamic adjustment mechanism for bald eagle flock size is introduced. In the initial stage of algorithm execution, the flock size is only 75% of its maximum capacity, with the remaining 25% reserved for later breeding and mortality. During algorithm iteration, the following principles are followed:

[0030] (1) If the globally optimal vulture individual is being updated, and N>N min Then delete b bald eagles;

[0031] (2) If the globally optimal bald eagle individual is not updated, and N <N max If b increases, then b more vultures will be added;

[0032] (3) If the globally optimal vulture individual is continuously T b No updates are made each time, and N > N min If so, then delete 2b vultures.

[0033] Where, N max It is the maximum capacity of a bald eagle flock; N min This represents the minimum capacity of a bald eagle flock. The mortality strategy targets the worst-performing individual in the flock, while the breeding strategy involves creating new bald eagles from the best and second-best bald eagles using the golden ratio method. The specific formula is as follows:

[0034]

[0035] In the formula, x new This indicates the location where a new bald eagle is generated after the t-th iteration; φ represents the second-best position of the vulture flock in the t-th iteration; φ represents the golden ratio.

[0036] Step 2 includes the following steps:

[0037] S2.1: The output power of the photovoltaic system is changed by adjusting the duty cycle d. The duty cycle d is regarded as the decision variable of the optimization problem, and the maximization of the output power P of the photovoltaic system is the optimization objective. The optimization model is as follows:

[0038] maxP = f(d);

[0039] std min ≤d≤d max ;

[0040] In the formula, d min and d max , respectively, represent the upper and lower limits of the duty cycle; f(d) represents the fitness function corresponding to the duty cycle d.

[0041] S2.2: The duty cycle is personified as the position of a vulture, and the output power of the photovoltaic system is personified as the success rate of the vulture catching fish in its position. An N-dimensional vector D = [d1, d2, ..., d...] is used. N ] and F = [f(d1), f(d2), ..., f(d N )] represent the location of a vulture flock of size N and the success rate of the vulture flock in catching fish at their respective locations; where: d1, d2, ..., d N These represent the magnitudes of N different duty cycles; f(d1), f(d2), ..., f(d... N ) represent the fitness functions corresponding to N different duty cycles.

[0042] S2.3: Boundary control operation. When the duty cycle exceeds its upper and lower limits, the duty cycle needs to be reset. The specific operation is as follows:

[0043]

[0044] In the formula, d best f(d) represents the maximum value f(d) in the N-dimensional vector F. best The duty cycle is α, which is a random number between (0,1).

[0045] d i This represents the magnitude of the i-th duty cycle.

[0046] S2.4: Restart Operation. When the external environment changes, the PV characteristic curve of the photovoltaic system will also change. At this time, it is necessary to restart the MPPT control algorithm to re-track the global maximum power point. This invention sets the restart conditions based on the power conversion amount as follows:

[0047]

[0048] In the formula, P n P represents the output power of the photovoltaic system at the current moment. m ΔP represents the maximum power value; ΔP represents the power conversion amount.

[0049] The process of applying the ABES algorithm for photovoltaic maximum power point tracking is as follows:

[0050] Step 1: Initialize N sets of duty cycles, form the position D of the vulture flock, and set the parameters of the algorithm;

[0051] Step 2: Calculate the success vector F of all vultures catching fish at their locations;

[0052] Step 3: Perform breeding and mortality operations on the vulture flock according to the dynamic adjustment mechanism of the vulture flock size;

[0053] Step 4: Update the position D of the vulture flock according to the stage iterative formula selected by the ABES algorithm, and perform the operation in Step 3;

[0054] Step 5: Update the position D of the vulture flock according to the iterative formula of the ABES algorithm search phase, and perform the operation in Step 3;

[0055] Step 6: Update the position D of the vulture flock according to the iterative formula of the ABES algorithm during the dive phase, and perform the operation in Step 3;

[0056] Step 7: Repeat the iterative process from Step 4 to Step 6 until the termination condition is met, then stop the search and output d. best and f(d) best These are respectively used as the optimal solution of the optimization model and the maximum output power of the photovoltaic system;

[0057] Step 8: Determine if the restart conditions are met. If the power conversion is greater than the set value, return to Step 1. If not, maintain the output d. best and f(d) best )constant.

[0058] Step 3 includes the following steps:

[0059] S3.1: Using the Simulink simulation platform, a simulation model of a photovoltaic power generation system is built. This simulation model consists of a photovoltaic array, a Boost circuit, a load, and an MPPT controller. The load is set to be a resistive load.

[0060] S3.2: The ABES algorithm is written as an MPPT module using the S-Function function. The input of this module is the output current and voltage of the photovoltaic array, and the output is the duty cycle of the Boost circuit. The duty cycle of the Boost circuit is controlled according to the output of the ABES algorithm module. The duty cycle is input to the PWM pulse signal generation module to generate a PWM pulse signal. The equivalent external resistance of the photovoltaic array is adjusted through the Boost circuit so that the equivalent external resistance of the photovoltaic array matches the equivalent internal resistance of the photovoltaic array, and the photovoltaic power generation system maintains the state of maximum output power.

[0061] S3.3: Applying the ABES algorithm to perform photovoltaic maximum power point tracking in four typical scenarios.

[0062] This invention discloses a photovoltaic maximum power point tracking method based on an adaptive vulture search algorithm, with the following technical advantages:

[0063] 1) In step 1 of this invention, in order to address the problems of the BES algorithm in photovoltaic MPPT control, a Gaussian mixture adaptive walk strategy and a progressive dive adaptive switching strategy are applied in the algorithm selection stage and the dive stage, respectively, so as to achieve a balance between global search and local optimization capabilities, improve optimization accuracy, and prevent the algorithm from getting stuck in local optima while accelerating the convergence speed through the eagle flock size adjustment mechanism.

[0064] ①: By introducing a Gaussian mixture adaptive walking strategy, the distribution of individual vultures is adjusted according to the number of iterations, thereby achieving global search and local search in the early and late stages of the algorithm, respectively;

[0065] ②: By simulating the escape behavior of fish, a progressive dive adaptive switching strategy is introduced to better simulate real-world situations;

[0066] ③: A dynamic adjustment mechanism for the size of a bald eagle flock is proposed. The flock size is adaptively adjusted based on whether the globally optimal bald eagle individual in several consecutive generations is updated. The bald eagle flock size adjustment mechanism can avoid the algorithm getting stuck in local optima and reduce tracking time.

[0067] These improvements aim to enhance the performance of the traditional vulture search algorithm in photovoltaic maximum power point tracking applications, and are also applicable to solving other nonlinear optimization problems.

[0068] 2) The photovoltaic maximum power point tracking method of the present invention can achieve satisfactory results in terms of tracking speed and tracking accuracy. The present invention not only considers the effectiveness of the algorithm when the external environment remains unchanged, but also considers the effectiveness of the algorithm when the external environment changes. Therefore, photovoltaic maximum power point tracking can be performed in different typical scenarios, which can improve the power generation of photovoltaic systems. Attached Figure Description

[0069] Figure 1 This is a flowchart of a photovoltaic maximum power point tracking method based on the ABES algorithm in this invention.

[0070] Figure 2 This is a flowchart illustrating the modeling process for photovoltaic maximum power point tracking using the ABES algorithm in this invention.

[0071] Figure 3 This is the MPPT control circuit diagram for a photovoltaic system.

[0072] Figure 4 Irradiance diagrams for four different operating conditions.

[0073] Figure 5 The graphs show the PV characteristic curves under four different operating conditions.

[0074] Figure 6(a) shows the output power curves of the photovoltaic system under different algorithms in scenario 1.

[0075] Figure 6(b) shows the output power curves of the photovoltaic system under different algorithms in scenario 2.

[0076] Figure 7(a) shows the output power curves of the photovoltaic system under different algorithms in scenario 3.

[0077] Figure 7(b) shows the output power curves of the photovoltaic system under different algorithms in scenario 4. Detailed Implementation

[0078] The specific embodiments and working principles of the present invention will be further described in detail below with reference to the accompanying drawings.

[0079] like Figure 1 As shown, the photovoltaic maximum power point tracking method based on the ABES algorithm includes the following steps:

[0080] S1: Construct the ABES algorithm;

[0081] In the selection phase of the BES algorithm, a Gaussian mixture adaptive walk strategy is introduced. The iterative formula for the selection phase of the ABES algorithm is as follows:

[0082]

[0083]

[0084] In the formula, This indicates the optimal position of the vulture flock at the t-th iteration; This indicates the center position of the vulture flock at the t-th iteration; N(μ,σ) represents the position of the i-th bald eagle in the t-th iteration; 2 ) indicates that it follows a sequence with expected value of μ and variance of σ. 2 Gaussian distributed random variable; t max This indicates the maximum number of iterations.

[0085] A progressive dive adaptive switching strategy is introduced in the dive phase of the BES algorithm. The iterative formula for the dive phase of the ABES algorithm is as follows:

[0086]

[0087] γ=r2-p s

[0088] In the formula, H(·) represents the Herveside step function; Sgn(·) represents the sign function; x1(i) and y1(i) represent the position of the vulture in polar coordinates; c1 and c2 represent the intensity of the vulture's movement towards the optimal and center positions; r1 and r2 both represent random numbers between (0,1); p sThis represents the probability that the fish successfully escapes. Its value decreases non-linearly from 0.75 to 0 as the number of iterations increases, using an adaptive sinusoidal decay method. The specific formula is as follows:

[0089]

[0090] When γ < 0, the fish have a chance to escape successfully, and the eagle flock adopts a gradual dive motion; when γ > 0, the fish have no chance to escape successfully, and the eagle flock adopts a dive motion. Considering that the intensity of the eagles' movement weakens over time, this invention proposes a nonlinear adaptive attenuation factor that takes into account the dynamic nature of the eagles' movement intensity, the specific formula of which is as follows:

[0091]

[0092] In the formula, c min and c max These are the minimum and maximum values ​​of parameters c1 and c2 within the range [1,2], respectively; β is the contraction coefficient.

[0093] A dynamic adjustment mechanism for bald eagle flock size is introduced. In the initial stage of algorithm execution, the flock size is only 75% of its maximum capacity, with the remaining 25% reserved for later breeding and mortality. During algorithm iteration, the following principles are followed:

[0094] (1) If the globally optimal vulture individual is being updated, and N>N min If so, then delete b bald eagles.

[0095] (2) If the globally optimal bald eagle individual is not updated, and N <N max If b increases, then b more vultures will be added.

[0096] (3) If the globally optimal vulture individual is continuously T b No updates are made each time, and N > N min If so, then delete 2b vultures.

[0097] Where, N max It is the maximum capacity of a bald eagle flock; N min This represents the minimum capacity of a bald eagle flock. The mortality strategy targets the worst-performing individual in the flock, while the breeding strategy involves creating new bald eagles from the best and second-best bald eagles using the golden ratio method. The specific formula is as follows:

[0098]

[0099] In the formula, x new This indicates the location where a new bald eagle is generated after the t-th iteration; φ represents the second-best position of the vulture flock in the t-th iteration; φ represents the golden ratio.

[0100] S2: Apply the ABES algorithm to photovoltaic maximum power point tracking;

[0101] S21: By adjusting the duty cycle, the output power of the photovoltaic system is changed. The duty cycle d is regarded as the decision variable of the optimization problem, and maximizing the output power P of the photovoltaic system is the optimization objective. The optimization model is as follows:

[0102] maxP=f(d)

[0103] std min ≤d≤d max

[0104] In the formula, d min and d max These are the upper and lower limits of the duty cycle, respectively.

[0105] S22: The duty cycle is personified as the position of a vulture, and the output power of the photovoltaic system is personified as the success rate of the vulture catching fish in its position. An N-dimensional vector D = [d1, d2, ..., d3] is used. N ] and F = [f(d1), f(d2), ..., f(d N )] represent the location of a flock of vultures with a population size of N and the success rate of the vultures in catching fish at their respective locations;

[0106] S23: Boundary control operation. When the duty cycle exceeds its upper or lower limit, the duty cycle needs to be reset. The specific operation is as follows:

[0107]

[0108] In the formula: d best f(d) represents the maximum value f(d) in the N-dimensional vector F. best The duty cycle is α, which is a random number between (0,1).

[0109] S24: Restart operation. When the external environment changes, the PV characteristic curve of the photovoltaic system will also change. At this time, it is necessary to restart the MPPT control algorithm to re-track the global maximum power point. This invention sets the restart conditions based on the power conversion amount as follows:

[0110]

[0111] In the formula, P n P represents the output power of the photovoltaic system at the current moment. m This indicates the maximum power value.

[0112] Figure 2 The process of using the ABES algorithm to determine the maximum power point of photovoltaic power is as follows:

[0113] (1) Initialize N sets of duty cycles, form the position D of the vulture flock, and set the parameters of the algorithm;

[0114] (2) Calculate the success vector F of all vultures catching fish at their locations;

[0115] (3) Implement breeding and mortality operations for bald eagle flocks based on the dynamic adjustment mechanism of bald eagle flock size;

[0116] (4) Update the position D of the vulture flock according to the stage iterative formula selected by the ABES algorithm, and perform the operation of step (3);

[0117] (5) Update the position D of the vulture flock according to the iterative formula of the search phase of the ABES algorithm, and perform the operation of step (3);

[0118] (6) Update the position D of the vulture flock according to the iterative formula of the ABES algorithm during the dive phase, and perform the operation of step (3);

[0119] (7) Repeat the iterative process from (4) to (6) until the termination condition is met, then stop the search and output d. best and f(d) best These are respectively used as the optimal solution of the optimization model and the maximum output power of the photovoltaic system;

[0120] (8) Determine if the restart condition is met. If the power conversion is greater than the set value, return to step (1). If not, maintain the output d. best and f(d) best )constant.

[0121] S3: Based on step S2, build a simulation model of the photovoltaic power generation system and verify the tracking performance of the ABES algorithm constructed in step S1 under different scenarios;

[0122] S31: Using the Simulink simulation platform, a simulation model of a photovoltaic power generation system is built. This simulation model consists of a photovoltaic array, a Boost circuit, a load, and an MPPT controller. The load is set to be a resistive load.

[0123] Depend on Figure 3 It can be seen that the input and output currents in the Boost circuit are I. pv I o Input and output voltages are U pv U o The load resistance is R. L The duty cycle is d; if the equivalent external resistance of the photovoltaic array is R, then we can obtain:

[0124]

[0125] From the above formula, we can obtain that if the load resistance R LIf the duty cycle d is constant, the equivalent external resistance R of the photovoltaic array can be adjusted by controlling the size of the duty cycle d, so that the equivalent external resistance of the photovoltaic array matches the equivalent internal resistance of the photovoltaic array, and the photovoltaic power generation system maintains the state of maximum output power.

[0126] S32: The ABES algorithm is written as an MPPT module using the S-Function function. The input of this module is the output current and voltage of the photovoltaic array, and the output is the duty cycle of the Boost circuit. The duty cycle of the Boost circuit is controlled according to the output of the ABES algorithm module. The duty cycle is input to the PWM pulse signal generation module to generate a PWM pulse signal. The equivalent external resistance of the photovoltaic array is adjusted through the Boost circuit so that the equivalent external resistance of the photovoltaic array matches the equivalent internal resistance of the photovoltaic array, and the photovoltaic power generation system maintains the state of maximum output power.

[0127] S33: Applying the ABES algorithm to perform photovoltaic maximum power point tracking in four typical scenarios;

[0128] A photovoltaic (PV) array is composed of PV modules connected in series and parallel. In actual operation, PV arrays are susceptible to localized shading, leading to multi-peak phenomena. To more effectively simulate the impact of localized shading on PV arrays, this invention uses a PV array composed of four PV modules connected in series as an example, setting irradiance levels under four different operating conditions as follows: Figure 4 As shown, the photovoltaic array is subjected to a temperature of 25℃, where the photovoltaic module PV 1-4 The shaded levels represent different levels of shading.

[0129] Figure 5 As shown Figure 4 The PV characteristic curves under four operating conditions are shown. Under condition 1, the PV characteristic curve exhibits a single peak, while under the other three conditions, the PV characteristic curves exhibit multi-peak phenomena. The red circles represent the global maximum power point under different operating conditions. It can be seen that under partial shading conditions, the PV characteristic curve will show multiple local extrema, and the location of the global maximum power point changes with the shading conditions. If the traditional MPPT algorithm is used, it is highly likely to track the local maximum power point, leading to a decrease in the output efficiency of the photovoltaic system. The heuristic intelligent algorithm has a strong global search capability and can track the global maximum power point better.

[0130] Example:

[0131] This implementation uses the Simulink simulation platform to build a photovoltaic MPPT control system based on a Boost circuit. Its circuit model is as follows: Figure 3 As shown. The circuit simulation parameters are set as follows: C1 = 0.01mF, C2 = 0.4676mF, L = 1.1478mH, R L=53Ω, the photovoltaic array is composed of 4 photovoltaic modules connected in series, and its detailed parameters are shown in Table 1.

[0132] Table 1 Component parameters of photovoltaic array

[0133]

[0134] In the Simulink environment, the ABES, BES, PSO, and DE-GWO algorithms were written as corresponding MPPT algorithm modules using the S-Function function. The population size of the ABES, BES, and PSO algorithms is 8, while the population size of the DE-GWO algorithm's secondary layer is set to 9. The initial duty cycle values ​​of each algorithm are uniformly distributed within their search space. The detailed parameter settings for the ABES algorithm and the three existing algorithms are shown in Table 2.

[0135] Table 2 shows the detailed parameter settings for the ABES algorithm and three existing algorithms.

[0136]

[0137] Note: w is the inertia weight in the PSO algorithm, and s1 and s2 are the learning factors in the PSO algorithm; F max and F min , where are the upper and lower limits of the scaling factor in the DE-GWO algorithm, and M is the number of gray wolves in the DE-GWO algorithm.

[0138] To verify the superiority of the ABES algorithm, maximum power point tracking was compared using the ABES, BES, PSO, and DE-GWO algorithms in four typical scenarios. The light intensity and temperature received by the photovoltaic modules in the four typical scenarios are shown in Table 3. The simulation duration for scenarios 1 and 2 was set to 2 seconds; the simulation duration for scenarios 3 and 4 was set to 4 seconds. At the 2nd second of the simulation time, the light intensity or temperature received by the photovoltaic modules changed.

[0139] Table 34 shows the light intensity and temperature received by photovoltaic modules under typical scenarios.

[0140]

[0141]

[0142] Figures 6(a) and 6(b) show the output power curves of the photovoltaic system under different algorithms in scenarios 1 and 2, respectively. The thick solid line represents the tracking result of the ABES algorithm, while the thin solid line, dotted line, and dashed line represent the tracking results of the BES, PSO, and DE-GWO algorithms, respectively. It can be seen that in scenarios 1 and 2, the ABES algorithm can track the vicinity of the global maximum power point earlier, and the ABES and DE-GWO algorithms are accompanied by smaller power fluctuations in this process; the BES and PSO algorithms have slower tracking speeds and are accompanied by larger power fluctuations in this process.

[0143] Figures 7(a) and 7(b) show the output power curves of the photovoltaic system under different algorithms in scenarios 3 and 4, respectively. The markings on each curve are consistent with those in Figures 6(a) and 6(b). The red dashed lines represent the moments when the light intensity or temperature changes. The following phenomena can be observed:

[0144] (1) In the 0-2s period of scenarios 3 and 4, the ABES algorithm and the three existing algorithms can all track the vicinity of the global maximum power point; the power fluctuations of the BES algorithm and PSO algorithm are relatively large in the early search process, and the BES algorithm has the slowest tracking speed; the power fluctuations of the DE-GWO algorithm are smaller than those of the BES algorithm and PSO algorithm in the early search process, but its tracking speed is slightly slower than that of the ABES algorithm; while the ABES algorithm can track the vicinity of the global maximum power point earlier.

[0145] (2) In the 2-4s period of scenario 3 and scenario 4, the ABES algorithm and the DE-GWO algorithm have the ability to respond to changes in the external environment more quickly, can restart the algorithm in time and quickly track the new global maximum power point, and can converge to a stable value after a small amount of power fluctuation during this process.

[0146] The above phenomena indicate that the ABES algorithm can effectively accelerate its convergence speed and solve the problem of large power fluctuations during the search process.

[0147] To more objectively analyze the performance metrics of the ABES algorithm and the three existing algorithms, the ABES algorithm and the three existing algorithms were run independently 15 times in four typical scenarios, and statistical comparisons were performed. The statistical results of the performance metrics of different algorithms in the four typical scenarios are shown in Table 4. The following phenomena can be observed:

[0148] (1) In four typical scenarios, the average and maximum values ​​of the optimization values ​​of the ABES algorithm and the existing three types of algorithms are close to the theoretical maximum power value.

[0149] (2) In the four typical scenarios, the ABES algorithm and the DE-GWO algorithm have high tracking efficiency. In scenario 1, the average tracking efficiency of the ABES algorithm and the DE-GWO algorithm is 99.95% and 99.90%, respectively, and the maximum is 99.97% and 99.96%, respectively. The average tracking efficiency of the BES algorithm and the PSO algorithm is 99.24% and 99.12%, respectively, and the maximum is 99.94% and 99.88%, respectively.

[0150] (3) In the four typical scenarios, the average and median energy efficiency of the ABES algorithm are higher than the average and median energy efficiency of the three existing algorithms. The difference between the maximum and minimum energy efficiency of the ABES algorithm is small. In scenario 2, the difference between the maximum and minimum energy efficiency of the ABES algorithm is 1.45%, while the differences between the maximum and minimum energy efficiency of the BES, PSO and DE-GWO algorithms are 6.82%, 8.19% and 6.69%, respectively.

[0151] The above phenomena indicate that the ABES algorithm and the three existing algorithms can successfully track the vicinity of the global maximum power point. On the other hand, the ABES algorithm can better improve the optimization accuracy and has better steady-state performance, which can improve the power generation of the photovoltaic system to a certain extent.

[0152] Table 4. Statistical results of different algorithm performance metrics in four typical scenarios.

[0153]

[0154] As can be seen from the results of the above embodiments, when applying the photovoltaic maximum power point tracking method based on the ABES algorithm of the present invention to perform maximum power point tracking in four typical scenarios, the photovoltaic maximum power point tracking method based on the ABES algorithm of the present invention has a faster convergence speed and smaller power fluctuation, and can avoid getting trapped in local optima under local shading. The photovoltaic maximum power point tracking method based on the ABES algorithm of the present invention can track the global maximum power point with high accuracy and stable performance in four typical scenarios, which can improve the power generation of the photovoltaic system to a certain extent.

Claims

1. A photovoltaic maximum power point tracking method based on an adaptive vulture search algorithm, characterized in that... Includes the following steps: Step 1: Based on the BES algorithm, introduce the Gaussian mixture adaptive walk strategy, the progressive dive adaptive switching strategy, and the eagle flock size adjustment mechanism to construct the ABES algorithm; Step 2: Apply the ABES algorithm constructed in Step 1 to photovoltaic maximum power point tracking; Step 3: Based on Step 2, build a simulation model of the photovoltaic power generation system to verify the tracking performance of the ABES algorithm constructed in Step 1 under different scenarios; In step 1, a Gaussian mixture adaptive walk strategy is introduced in the BES algorithm selection phase. The iterative formula for the ABES algorithm selection phase is as follows: ； ； In the formula: Indicates the first t At the +1st iteration, the... i The location of the vulture; Indicates the first t The optimal position for the eagle flock in the next iteration; Indicates the first t The center position of the eagle flock at the next iteration; Indicates the first t During the nth iteration i The location of the vulture; N ( μ , σ 2 ) indicates compliance with expectations. μ The variance is σ 2 Gaussian distributed random variable; Indicates the maximum number of iterations; The variance is expressed as Weights of the Gaussian distribution; The variance is expressed as Weights of the Gaussian distribution; Indicates the number of iterations; This represents the standard deviation of the first Gaussian distribution; This represents the standard deviation of the second Gaussian distribution; A progressive dive adaptive switching strategy is introduced in the dive phase of the BES algorithm. The iterative formula for the dive phase of the ABES algorithm is as follows: ； In the formula, H (·) represents the Herveside step function; Sgn (·) denotes a sign function; x 1( i )and y 1( i () represents the position of the vulture in polar coordinates; c 1 and c 2 indicates the intensity of the vulture's movement towards the optimal and central position; r 1 and r 2 represents a random number between (0, 1); This indicates that the fish has escaped. This indicates the probability that the fish successfully escaped. p s The probability of a fish successfully escaping is expressed by the following formula: ； when γ< At 0:00, the fish have a chance to escape successfully, and the flock of vultures adopts a gradual dive. when γ> At 0:00, the fish had no chance to escape, and the flock of vultures swooped down. A nonlinear adaptive attenuation factor considering the dynamics of the vulture's movement intensity is proposed, and the specific formula is as follows: ； In the formula, c min and c max Parameters c 1 and c 2. The minimum and maximum values ​​within the range [1,2]. β It is the coefficient of shrinkage; A dynamic adjustment mechanism for the size of the bald eagle flock is introduced. In the early stage of algorithm execution, the size of the bald eagle flock is only 75% of the maximum capacity, and the remaining 25% is prepared for the later breeding and death of the bald eagle flock. The following principles should be followed during the algorithm iteration process: (1) If the globally optimal vulture individual is being updated, and N > N min , then delete b A vulture; (2) If the globally optimal bald eagle individual is not updated, and N < N max Then increase b A vulture; (3) If the globally optimal vulture individual is continuous T b It never updates, and N > N min Then delete 2 b A vulture; in, N max This is the maximum capacity of a flock of vultures. N min This represents the minimum capacity of a bald eagle flock; the mortality strategy targets the worst-performing individual in the flock, while the breeding strategy involves generating new bald eagles from the best and second-best bald eagles using the golden ratio method, as shown in the following formula: ； In the formula, x new Indicates the first t The location of the new bald eagle is generated after the next iteration; Indicates the first t The second-best position of the eagle flock in the next iteration; This represents the golden ratio.

2. The photovoltaic maximum power point tracking method based on the adaptive vulture search algorithm according to claim 1, characterized in that: Step 2 includes the following steps: S2.1: By adjusting the duty cycle d This, in turn, changes the output power and duty cycle of the photovoltaic system. d The output power of the photovoltaic system is considered a decision variable in an optimization problem. P With maximization as the optimization objective, the optimization model is as follows: ； ； In the formula, d min and d max These are the upper and lower limits of the duty cycle, respectively. Indicates duty cycle as d The corresponding fitness function; S2.2: The duty cycle is personified as the position of a vulture, and the photovoltaic system's output power is personified as the success rate of the vulture catching fish in its position. N dimensional vector D =[ d 1, d 2,…, d N ]and F =[ f ( d 1), f ( d 2),…, f ( d N )] represent population size respectively. N The location of the vulture flock and the success rate of the vulture flock in fishing in their respective locations; among which: d 1, d 2,…, d N They represent N The size of different duty cycles; f ( d 1), f ( d 2),…, f ( d N ) respectively represent N Fitness functions corresponding to different duty cycles; S2.3: Boundary control operation. When the duty cycle exceeds its upper and lower limits, the duty cycle needs to be reset. The specific operation is as follows: ； In the formula, d best express N dimensional vector F maximum value f ( d best The corresponding duty cycle; α It is a random number between (0,1); Indicates the first i The size of the duty cycle; S2.4: Restart Operation. When the external environment changes, the PV characteristic curve of the photovoltaic system will also change. At this time, the MPPT control algorithm is restarted to re-track the global maximum power point. The restart conditions are set based on the power conversion amount as follows: ； In the formula, P n This indicates the current output power of the photovoltaic system; P m Indicates the maximum power value; This indicates the amount of power conversion.

3. The photovoltaic maximum power point tracking method based on the adaptive vulture search algorithm according to claim 2, characterized in that: The process of applying the ABES algorithm for photovoltaic maximum power point tracking is as follows: Step 1: Initialization N Group duty cycle, forming the location of the vulture flock D Set the parameters of the algorithm; Step 2: Calculate the success vector of all vultures catching fish at their locations. F ; Step 3: Perform breeding and mortality operations on the vulture flock according to the dynamic adjustment mechanism of the vulture flock size; Step 4: Update the position of the vulture flock according to the stage iterative formula selected by the ABES algorithm. D Then proceed with Step 3; Step 5: Update the position of the vulture flock according to the iterative formula of the ABES algorithm search phase. D Then proceed with Step 3; Step 6: Update the position of the vulture flock according to the iterative formula of the ABES algorithm during the dive phase. D Then proceed with Step 3; Step 7: Repeat the iterative process from Step 4 to Step 6 until the termination condition is met, then stop the search and output the results. d best and f ( d best These are respectively used as the optimal solution of the optimization model and the maximum output power of the photovoltaic system; Step 8: Determine if the restart conditions are met. If the power conversion is greater than the set value, return to Step 1; otherwise, maintain the output. d best and f ( d best )constant.

4. The photovoltaic maximum power point tracking method based on the adaptive vulture search algorithm according to claim 1, characterized in that: Step 3 includes the following steps: S3.1: Using the Simulink simulation platform, a simulation model of a photovoltaic power generation system is built. This simulation model consists of a photovoltaic array, a Boost circuit, a load, and an MPPT controller. The load is set to be a resistive load. S3.2: The ABES algorithm is written as an MPPT module using the S-Function function. The input of this module is the output current and voltage of the photovoltaic array, and the output is the duty cycle of the Boost circuit. The duty cycle of the Boost circuit is controlled according to the output of the ABES algorithm module. The duty cycle is input to the PWM pulse signal generation module to generate a PWM pulse signal. The equivalent external resistance of the photovoltaic array is adjusted through the Boost circuit so that the equivalent external resistance of the photovoltaic array matches the equivalent internal resistance of the photovoltaic array, and the photovoltaic power generation system maintains the state of maximum output power. S3.3: Applying the ABES algorithm to perform photovoltaic maximum power point tracking in four typical scenarios.

Images