Photovoltaic system maximum power point tracking method

By combining the inertia factor and acceleration coefficient update formula, based on the k-mean clustering and differential evolution algorithm, the particle swarm algorithm falls into the local optimal problem in the maximum power point tracking of photovoltaics, and improves the maximum power point tracking efficiency of photovoltaic systems and Dynamic stability.

CN118963483APending Publication Date: 2024-11-15HUANENG HUALIANGTING HYDROPOWER CO LTD
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202410944673.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-15
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

Existing particle swarm algorithms are prone to fall into local optimality in photovoltaic maximum power point tracking, resulting in reduced efficiency and lack of dynamic stability.

Method used

Combining the inertia factor and acceleration coefficient update formula, particle clustering is performed based on the k-mean clustering algorithm, differential evolution algorithm is used for cross-selecting, and algorithm restart conditions are set to prevent changes in the maximum power point.

Benefits of technology

It improves the algorithm's global search capability and convergence speed, enhances dynamic stability, avoids the reduction of maximum power point tracking efficiency, and ensures that the photovoltaic system can respond to power changes in a timely manner.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118963483A_ABST
    Figure CN118963483A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of photovoltaic power generation, and discloses a photovoltaic system maximum power point tracking method, which comprises the following steps of: determining an individual optimal value and a global optimal value; updating an inertia factor and an acceleration coefficient, and updating the position and the speed of each particle according to a particle iteration formula; carrying out clustering division on the particle swarm based on a k-means clustering algorithm; the particles are crossed and selected based on a differential evolution algorithm, the final particles are updated, and an individual optimal value and a global optimal value are calculated; judging whether the current iteration number or the distance between particles meets an algorithm termination condition or not, if yes, outputting the maximum power value according to the global optimal value, and if not, returning to the second step; and judging whether the algorithm is restarted or not according to the power variation, and if the algorithm does not need to be restarted, outputting a global optimal value. The maximum power point of the photovoltaic system is tracked based on the particle swarm optimization and in combination with the differential evolution algorithm, the global search capability and dynamic stability of the algorithm are improved, and the power generation efficiency of the photovoltaic system is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field of photovoltaic power generation, and more specifically, to a maximum power point tracking method for a photovoltaic system. Background Art

[0002] Solar energy reserves are huge, and its development is not geographically restricted and pollution-free. Therefore, solar power generation technology has important research value and broad application prospects. Although solar energy, as a new energy source with huge potential, has been paid attention to and valued by countries around the world, and the photovoltaic industry has received strong support from my country, there are still many technical problems in the photovoltaic industry. The energy conversion materials used in the photovoltaic industry are mainly semiconductors such as monocrystalline silicon, and their energy conversion efficiency is only about 20%, which is very low. This has seriously restricted the development of the photovoltaic industry. The maximum power tracking technology to improve the efficiency of photovoltaic power generation has become a research hotspot in the current photovoltaic industry.

[0003] Among the existing multi-peak MPPT algorithms, the particle swarm (PSO) algorithm is widely used by researchers due to its advantages of low complexity, fast convergence speed and simple control. There are two parts in the particle swarm velocity update formula: self-cognition component and social cognition component. These two components have an important influence on the movement distance and direction of particles. This structure leads to too fast information exchange between particles and a single direction. As the iteration proceeds, the diversity of particles is quickly lost and the population quickly converges to a smaller space, which makes the PSO algorithm easy to fall into the local optimum in the optimization problem.

[0004] Therefore, how to improve the global search capability of the particle swarm algorithm, prevent it from falling into the local optimum, and enhance the dynamic stability of the algorithm are technical problems that need to be solved. Summary of the invention

[0005] The present invention provides a photovoltaic system maximum power point tracking method to solve the technical problem that photovoltaic maximum power point tracking in the prior art is prone to fall into local optimum, resulting in reduced efficiency of the photovoltaic system. The method comprises:

[0006] S101, initializing particle swarm parameters;

[0007] S102, taking the output power of the photovoltaic system as the fitness function, calculating the fitness function value of each particle, and determining the individual optimal value and the global optimal value;

[0008] S103, updating the inertia factor and acceleration coefficient, and updating the position and velocity of each particle according to the particle iteration formula, where the particle position is the duty cycle of the photovoltaic system, and the particle velocity is the change in the duty cycle of the photovoltaic system;

[0009] S104, clustering the particle swarm based on the k-means clustering algorithm;

[0010] S105, crossover and selection of particles based on differential evolution algorithm, update final particles, and calculate individual optimal values ​​and global optimal values;

[0011] S106, judging whether the current number of iterations or the distance between particles meets the algorithm termination condition, if so, outputting the maximum power value according to the global optimal value, if not, returning to step S102;

[0012] S107 , judging whether to restart the algorithm according to the power change amount, and if the algorithm does not need to be restarted, outputting a global optimal value.

[0013] Furthermore, the inertia factor and acceleration coefficient are:

[0014] w=(w min -w max )(K max -k) / K max +w max

[0015]

[0016] Among them, w is the inertia factor, w min is the initial inertia factor, w max is the maximum inertia factor, K max is the maximum number of iterations, k is the current number of iterations, c1 and c2 are acceleration coefficients, B1 is the first preset range coefficient, B2 is the second preset range coefficient, and B1<B2.

[0017] Furthermore, clustering the particle swarm based on the k-means clustering algorithm includes:

[0018] Randomly select n particle cluster centers and calculate the distance from each particle to the cluster center;

[0019] Cluster the particles to the nearest cluster center to form m particle clusters;

[0020] Calculate the cluster center of each particle cluster according to the mean value of particles in the particle cluster, and update the particle cluster center according to the particle cluster cluster center;

[0021] If the number of clustering algorithm iterations reaches the preset number of re-clustering times, the particle clustering center is re-selected for re-clustering;

[0022] If the particle cluster center no longer changes or reaches the maximum number of iterations preset by the clustering algorithm, the particles are updated according to the cluster center.

[0023] Furthermore, the preset re-clustering times calculation formula is:

[0024]

[0025] Among them, T is the preset number of re-clustering times, Q max is the preset maximum number of iterations, and q is the current number of iterations.

[0026] Furthermore, the crossover and selection of particles based on the differential evolution algorithm includes:

[0027] Set the crossover probability and set the particle velocity calculation formula according to the crossover probability;

[0028] According to the particle fitness function value, set the particle position calculation formula.

[0029] Furthermore, the particle velocity calculation formula is:

[0030]

[0031] in, and is the particle velocity of the i-th particle at the k+1th and kth iterations, is the particle position of the i-th particle at the k-th iteration, r1 is the first random number, G best is the global optimal value, and CR is the crossover probability.

[0032] Furthermore, the particle position calculation formula is:

[0033]

[0034] in, is the particle position of the i-th particle at the k+1th iteration, is the transition position variable, and for and The corresponding photovoltaic system power.

[0035] Furthermore, the termination condition of the algorithm is:

[0036] Get the current number of iterations and determine whether the current number of iterations reaches the preset maximum number of iterations. If so, terminate the algorithm.

[0037] If not, it is determined whether the distance between particles is less than a first preset threshold. If so, the algorithm is terminated.

[0038] Further, judging whether to restart the algorithm according to the power change amount includes:

[0039] The maximum output power value and the current power of the photovoltaic system are obtained, the power change of the photovoltaic system is calculated, and it is determined whether the power change is greater than a second preset threshold. If so, the algorithm is restarted.

[0040] Furthermore, the power change calculation formula is:

[0041]

[0042] Where ΔP is the power change, P a is the current power of the photovoltaic system, P max is the maximum output power value.

[0043] The beneficial effects of the present invention are:

[0044] By applying the above technical scheme, the present invention performs photovoltaic maximum power point tracking based on a particle swarm algorithm in combination with a differential evolution algorithm. By updating the calculation parameters of the particle swarm algorithm, the algorithm always performs at its best during maximum power point tracking. By performing k-means clustering of the particle swarm before differential evolution, the convergence speed of the algorithm is improved, and the dynamic stability is enhanced. At the same time, by setting the algorithm restart condition, the maximum power point is prevented from changing due to changes in external conditions, which causes the photovoltaic system to be unable to respond to the change in the maximum power point in a timely manner, thereby greatly improving the maximum power point tracking efficiency of the photovoltaic system. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0046] Figure 1 A logic block diagram of a photovoltaic system maximum power point tracking method proposed in an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0047] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0048] The present application provides a method for tracking the maximum power point of a photovoltaic system. Figure 1 As shown, the method comprises the following steps:

[0049] S101, initializing particle swarm parameters;

[0050] S102, taking the output power of the photovoltaic system as the fitness function, calculating the fitness function value of each particle, and determining the individual optimal value and the global optimal value;

[0051] S103, updating the inertia factor and acceleration coefficient, and updating the position and velocity of each particle according to the particle iteration formula, where the particle position is the duty cycle of the photovoltaic system, and the particle velocity is the change in the duty cycle of the photovoltaic system;

[0052] S104, clustering the particle swarm based on the k-means clustering algorithm;

[0053] S105, crossover and selection of particles based on differential evolution algorithm, update final particles, and calculate individual optimal values ​​and global optimal values;

[0054] S106, judging whether the current number of iterations or the distance between particles meets the algorithm termination condition, if so, outputting the maximum power value according to the global optimal value, if not, returning to step S102;

[0055] S107 , judging whether to restart the algorithm according to the power change amount, and if the algorithm does not need to be restarted, outputting a global optimal value.

[0056] In some embodiments of the present application, the inertia factor and acceleration coefficient are:

[0057] w=(w min -w max )(K max -k) / K max +w max

[0058]

[0059] Among them, w is the inertia factor, w min is the initial inertia factor, w max is the maximum inertia factor, K max is the maximum number of iterations, k is the current number of iterations, c1 and c2 are acceleration coefficients, B1 is the first preset range coefficient, B2 is the second preset range coefficient, and B1<B2.

[0060] In this embodiment, the first preset range coefficient is set to 0.5, and the second preset range coefficient is set to 2.5. The inertia factors w and c1 decrease linearly during the particle iteration process. In the early stage of the algorithm, the inertia factor and c1 are large, and the population focuses on exploring the unknown space to avoid premature convergence. In the later stage of the algorithm, the inertia factor becomes smaller and c2 is larger. The population strengthens the search for the local range where the optimal individual is located, improves the accuracy of the algorithm, improves the convergence speed of the algorithm, avoids the algorithm from falling into the local optimum, and enables the particle swarm algorithm to perform at its best performance.

[0061] In some embodiments of the present application, the clustering of particle swarms based on the k-means clustering algorithm includes: randomly selecting n particle clustering centers and calculating the distance from each particle to the clustering center; clustering the particles to the nearest clustering center to form m particle clusters; calculating the clustering center of each particle cluster according to the mean of the particles in the particle cluster, and updating the particle clustering center according to the particle cluster clustering center; if the number of clustering algorithm iterations reaches a preset number of re-clustering times, reselecting the particle clustering center and re-clustering; if the particle clustering center no longer changes or reaches the preset maximum number of iterations of the clustering algorithm, updating the particles according to the clustering center.

[0062] In this embodiment, before the particles are differentially evolved, the k-means clustering algorithm is used to cluster the particles, and the Euclidean distance is used as the similarity evaluation index. The closer the distance between two particles, the greater the similarity. The number of initial cluster centers is selected according to the number of empirical peaks of the multi-peak power curve, and then the distance from each particle to the center is calculated, and the particle is assigned to the center point closest to it, thereby improving the convergence speed of the subsequent differential evolution algorithm.

[0063] In some embodiments of the present application, the preset re-clustering times calculation formula is:

[0064]

[0065] Among them, T is the preset number of re-clustering times, Q max is the preset maximum number of iterations, and q is the current number of iterations.

[0066] In this embodiment, in order to reduce the impact of the initial cluster center on the results, the algorithm re-clusters every T iterations. In the early stage of the algorithm, the value of T is set to be small to enhance the development capability of the algorithm; in the later stage of the algorithm, the value of T is set to be large to enhance the exploration capability of the algorithm.

[0067] In some embodiments of the present application, the crossover and selection of particles based on the differential evolution algorithm includes:

[0068] Set the crossover probability and set the particle velocity calculation formula according to the crossover probability;

[0069] According to the particle fitness function value, set the particle position calculation formula.

[0070] In some embodiments of the present application, the particle velocity calculation formula is:

[0071]

[0072] in, and is the particle velocity of the i-th particle at the k+1th and kth iterations, is the particle position of the i-th particle at the k-th iteration, r1 is the first random number, G best is the global optimal value, and CR is the crossover probability.

[0073] In this embodiment, the speed of the next generation of particles is determined according to the crossover probability, the competition relationship between particles is strengthened, the algorithm is prevented from falling into the local optimum, and the convergence speed is further improved.

[0074] In some embodiments of the present application, the particle position calculation formula is:

[0075]

[0076] in, is the particle position of the i-th particle at the k+1th iteration, is the transition position variable, and for and The corresponding photovoltaic system power.

[0077] In this embodiment, the position of the final particle is determined by selecting particles with higher fitness function values ​​based on the differential evolution algorithm, which speeds up the process of particles moving toward the global optimal value, makes the search area of ​​the particles more concentrated, and saves search time.

[0078] In some embodiments of the present application, the algorithm termination condition is:

[0079] Get the current number of iterations and determine whether the current number of iterations reaches the preset maximum number of iterations. If so, terminate the algorithm.

[0080] If not, it is determined whether the distance between particles is less than a first preset threshold. If so, the algorithm is terminated.

[0081] In this embodiment, the particle swarm algorithm can make the particles continuously approach the global optimal value. When the maximum distance between all adjacent particles reaches the preset threshold, no iteration is performed to prevent the maximum power jitter and accelerate the convergence speed. In this embodiment, the distance between particles is the Euclidean distance, the preset threshold can be set to 0.2, and the preset maximum number of iterations can be set to 100.

[0082] In some embodiments of the present application, judging whether to restart the algorithm according to the power change amount includes:

[0083] The maximum output power value and the current power of the photovoltaic system are obtained, the power change of the photovoltaic system is calculated, and it is determined whether the power change is greater than a second preset threshold. If so, the algorithm is restarted.

[0084] In some embodiments of the present application, the power change calculation formula is:

[0085]

[0086] Where ΔP is the power change, P a is the current power of the photovoltaic system, P max is the maximum output power value.

[0087] In this embodiment, after the maximum power point is output, when the external environment of the photovoltaic system changes significantly, the maximum power point will change. Therefore, the power change calculation formula of this embodiment is used to detect the change of the maximum power point. When ΔP>0.1, it indicates that the maximum power point has changed at this time, and the algorithm needs to be restarted to ensure that the photovoltaic system can track the maximum power point in time and enhance dynamic stability.

[0088] In order to further explain the technical idea of ​​the present invention, the technical solution of the present invention is now described in combination with specific application scenarios.

[0089] By applying the above technical solution, after the particle swarm parameters are initialized, the output power of the photovoltaic system is used as the fitness function to determine the individual optimal value and the global optimal value, and then the inertia factor and the acceleration coefficient are updated to improve the global search ability and convergence speed of the algorithm, the particle position is set to the duty cycle of the photovoltaic system, the particle speed is the change in the duty cycle of the photovoltaic system, the particle position is calculated according to the updated iterative formula, the duty cycle is updated, the particle swarm is clustered based on the k-means clustering algorithm, and high-quality particles are provided for the subsequent differential evolution algorithm. The crossover and selection operations in the differential algorithm are introduced into the algorithm of this embodiment to prevent the algorithm from falling into the local optimum, the final particles are updated, and the individual optimal value and the global optimal value are calculated; the termination condition of the algorithm is linked to the maximum distance between adjacent particles, so that the algorithm can be terminated in advance, the convergence speed is improved, and the maximum power point is prevented from oscillating; whether to restart the algorithm is determined according to the power change, if the algorithm does not need to be restarted, the global optimal value is output, if it needs to be restarted, the step of initializing the particle swarm parameters is returned to restart the algorithm.

[0090] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present invention can be implemented by hardware, or by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present invention can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each implementation scenario of the present invention.

[0091] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A photovoltaic system maximum power point tracking method, characterized in that: The method comprises: S101, initializing particle swarm parameters; S102, taking the output power of the photovoltaic system as the fitness function, calculating the fitness function value of each particle, and determining the individual optimal value and the global optimal value; S103, updating the inertia factor and acceleration coefficient, and updating the position and velocity of each particle according to the particle iteration formula, where the particle position is the duty cycle of the photovoltaic system, and the particle velocity is the change in the duty cycle of the photovoltaic system; S104, clustering the particle swarm based on the k-means clustering algorithm; S105, crossover and selection of particles based on differential evolution algorithm, update final particles, and calculate individual optimal values ​​and global optimal values; S106, judging whether the current number of iterations or the distance between particles meets the algorithm termination condition, if so, outputting the maximum power value according to the global optimal value, if not, returning to step S102; S107 , judging whether to restart the algorithm according to the power change amount, and if the algorithm does not need to be restarted, outputting a global optimal value.

2. The photovoltaic system maximum power point tracking method according to claim 1, characterized in that: The inertia factor and acceleration coefficient are: w=(w min -w max )(K max -k) / K max +w max Among them, w is the inertia factor, w min is the initial inertia factor, w max is the maximum inertia factor, K max is the maximum number of iterations, k is the current number of iterations, c1 and c2 are acceleration coefficients, B1 is the first preset range coefficient, B2 is the second preset range coefficient, and B1<B2.

3. The photovoltaic system maximum power point tracking method according to claim 1, characterized in that: The clustering of the particle swarm based on the k-means clustering algorithm includes: Randomly select n particle cluster centers and calculate the distance from each particle to the cluster center; Cluster the particles to the nearest cluster center to form m particle clusters; Calculate the cluster center of each particle cluster according to the mean value of particles in the particle cluster, and update the particle cluster center according to the particle cluster cluster center; If the number of clustering algorithm iterations reaches the preset number of re-clustering times, the particle clustering center is re-selected for re-clustering; If the particle cluster center no longer changes or reaches the maximum number of iterations preset by the clustering algorithm, the particles are updated according to the cluster center.

4. The photovoltaic system maximum power point tracking method according to claim 3, characterized in that: The preset re-clustering times calculation formula is: Among them, T is the preset number of re-clustering times, Q max is the preset maximum number of iterations, and q is the current number of iterations.

5. The photovoltaic system maximum power point tracking method according to claim 1, characterized in that: The crossover and selection of particles based on the differential evolution algorithm includes: Set the crossover probability and set the particle velocity calculation formula according to the crossover probability; According to the particle fitness function value, set the particle position calculation formula.

6. The photovoltaic system maximum power point tracking method according to claim 5, characterized in that: The particle velocity calculation formula is: in, and is the particle velocity of the i-th particle at the k+1th and kth iterations, is the particle position of the i-th particle at the k-th iteration, r1 is the first random number, G best is the global optimal value, and CR is the crossover probability.

7. The photovoltaic system maximum power point tracking method according to claim 6, characterized in that: The particle position calculation formula is: in, is the particle position of the i-th particle at the k+1th iteration, is the transition position variable, and for and The corresponding photovoltaic system power.

8. The photovoltaic system maximum power point tracking method according to claim 1, characterized in that: The termination condition of the algorithm is: Get the current number of iterations and determine whether the current number of iterations reaches the preset maximum number of iterations. If so, terminate the algorithm. If not, it is determined whether the distance between particles is less than a first preset threshold. If so, the algorithm is terminated.

9. The photovoltaic system maximum power point tracking method according to claim 1, characterized in that: The step of judging whether to restart the algorithm according to the power change comprises: The maximum output power value and the current power of the photovoltaic system are obtained, the power change of the photovoltaic system is calculated, and it is determined whether the power change is greater than a second preset threshold. If so, the algorithm is restarted.

10. The photovoltaic system maximum power point tracking method according to claim 9, characterized in that: The power change calculation formula is: Where ΔP is the power change, P a is the current power of the photovoltaic system, P max is the maximum output power value.

Citation Information

Cited By

  • Photovoltaic power station power generation monitoring system based on edge calculation

    CN119254143A