Photovoltaic array global maximum power point tracking method and system based on svm and feature mapping
By using SVM and feature mapping-based methods to extract feature points of the PV curve of a photovoltaic array and optimize the parameters of the support vector machine, combined with the adaptive perturbation observation method, the problems of high hardware cost, slow speed and poor universality in the existing technology are solved, and high-precision and fast maximum power point tracking is achieved.
Patent Information
- Application Number
- CN202211102186.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-09-09
- Publication Date
- 2026-02-24
- Estimated Expiration
- 2042-09-09
AI Technical Summary
Existing maximum power point tracking technology cannot effectively reduce hardware costs, improve tracking speed, and enhance versatility while maintaining accuracy.
A method based on support vector machine (SVM) and feature mapping is adopted. By modeling the PV curve of photovoltaic array under different shading conditions, the voltage, power and slope values of feature points are extracted, a training set is constructed and the SVM parameters are optimized. The maximum power point is tracked by combining the adaptive perturbation observation method.
While ensuring tracking accuracy, the tracking speed of the global maximum power point of the photovoltaic array is significantly improved, enhancing the versatility and tracking speed of the method.
Smart Images
Figure CN115657782B_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of maximum power point tracking, and particularly relates to a method and system for global maximum power point tracking of photovoltaic arrays based on SVM and feature mapping. Background Technology
[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.
[0003] In recent years, the installed capacity of renewable energy systems has experienced explosive growth, with photovoltaic (PV) power generation showing the most significant increase. PV system output exhibits pronounced nonlinear characteristics; when a PV array is shaded, its maximum power point (MPPT) will shift significantly. To improve the economics of PV power generation systems, maximum power point tracking (MPPT) technology needs optimization. Currently, numerous MPPT methods have been proposed, varying in terms of hardware cost, computational complexity, tracking speed, tracking accuracy, hardware implementation difficulty, and universality. Finding a better balance among these factors is the direction for optimizing MPPT methods.
[0004] Existing global maximum power point tracking (GMPT) algorithms can be divided into three main categories. The first category is swarm intelligence optimization algorithms. These methods can achieve accurate GMPT tracking without additional hardware; however, their relatively low convergence speed limits their industrial application prospects. The second category can be collectively referred to as two-stage methods. The first stage quickly locates the GMPT interval, and the second stage performs small-scale perturbations based on the first stage to achieve accurate GMPT tracking. Although this method has a high tracking speed and can accurately track GMPT under complex shading conditions, the scanning process of the PV curve is quite cumbersome, and high-performance digital controllers are required for high-speed tracking, resulting in low versatility. The third category is artificial intelligence algorithms. Their main idea is to establish a mapping relationship between meteorological parameters and the GMPT of the photovoltaic array, and then achieve rapid GMPT tracking under different operating conditions based on measured meteorological parameters. However, the irradiance and temperature of each component of the photovoltaic array are difficult to obtain in practical applications, and the accuracy is difficult to guarantee. Therefore, this type of algorithm has low versatility.
[0005] In summary, existing maximum power point tracking (MPPT) technologies cannot reduce the hardware cost, improve the tracking speed, or enhance the versatility of the MPT algorithm while maintaining MPT accuracy. Summary of the Invention
[0006] To overcome the shortcomings of the prior art, this invention provides a photovoltaic array global maximum power point tracking method and system based on SVM and feature mapping. Addressing the problem that existing global maximum power point tracking methods still cannot effectively balance hardware cost, tracking speed, tracking accuracy, and universality, this invention proposes a photovoltaic array global maximum power point tracking method based on SVM and feature mapping, which significantly improves the tracking speed of the global maximum power point of the photovoltaic array while ensuring tracking accuracy.
[0007] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions:
[0008] The first aspect of this invention provides a method for global maximum power point tracking of photovoltaic arrays based on SVM and feature mapping.
[0009] The global maximum power point tracking method for photovoltaic arrays based on SVM and feature mapping includes the following steps:
[0010] Model the photovoltaic array and collect the PV curves of the photovoltaic array under different shading conditions;
[0011] Feature points of the PV curve are selected, and the voltage, power, and slope values corresponding to the feature points are extracted.
[0012] A training set is constructed based on the extracted feature point voltage, power, and slope values. The support vector machine is then trained to obtain a well-trained support vector machine.
[0013] The parameters of the trained support vector machine are optimized to obtain the optimized support vector machine.
[0014] The output voltage and current of the photovoltaic array are sampled in real time to determine the type of shading of the photovoltaic array;
[0015] Based on the occlusion type, the maximum power point is tracked using an optimized support vector machine and an adaptive P&O algorithm.
[0016] The second aspect of the present invention provides a global maximum power point tracking system for photovoltaic arrays based on SVM and feature mapping.
[0017] A global maximum power point tracking system for photovoltaic arrays based on SVM and feature mapping includes:
[0018] The PV curve acquisition module is configured to: model the photovoltaic array and collect the PV curves of the photovoltaic array under different shading conditions;
[0019] The feature point selection module is configured to: select feature points of the PV curve and extract the voltage, power and slope values corresponding to the feature points;
[0020] The training module is configured to: construct a training set based on the extracted feature point voltage, power, and slope values, train the support vector machine, and obtain a trained support vector machine;
[0021] The optimization module is configured to optimize the parameters of the trained support vector machine to obtain an optimized support vector machine.
[0022] The shading type determination module is configured to: sample the output voltage and output current of the photovoltaic array in real time and determine the shading type of the photovoltaic array;
[0023] The maximum power point tracking module is configured to track the maximum power point based on an optimized support vector machine and an adaptive P&O algorithm, depending on the occlusion type.
[0024] A third aspect of the present invention provides a computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps of the photovoltaic array global maximum power point tracking method based on SVM and feature mapping as described in the first aspect of the present invention.
[0025] The fourth aspect of the present invention provides an electronic device including a memory, a processor, and a program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps in the photovoltaic array global maximum power point tracking method based on SVM and feature mapping as described in the first aspect of the present invention.
[0026] The above one or more technical solutions have the following beneficial effects:
[0027] 1. This invention addresses the problem that existing global maximum power point tracking methods still cannot effectively balance hardware cost, tracking speed, tracking accuracy, and universality. It proposes a global maximum power point tracking method for photovoltaic arrays based on SVM and feature mapping, which significantly improves the tracking speed of the global maximum power point of photovoltaic arrays while ensuring tracking accuracy.
[0028] 2. This invention optimizes the artificial intelligence algorithm, taking the shape characteristics of the PV curve of the photovoltaic array as the starting point, and achieves accurate and fast tracking of the maximum power point without the need for meteorological data, which significantly improves the versatility of the method.
[0029] 3. This invention establishes a mapping relationship between the shape features of the PV curve and the global maximum power point based on support vector machines. It achieves rapid localization of the global maximum power point by scanning several feature points on the PV curve. Based on the particle swarm optimization algorithm, the values of the support vector machine penalty optimization parameters and kernel function parameters are optimized to improve the prediction accuracy of the voltage at the global maximum power point and shorten the maximum power point tracking time. Combined with the adaptive perturbation observation method, high-precision global maximum power point tracking is further achieved.
[0030] 4. Unlike traditional global maximum power point tracking methods, this invention performs feature analysis on the shape characteristics of the photovoltaic array PV curve, which significantly improves the tracking speed.
[0031] Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description
[0032] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.
[0033] Figure 1 This is a flowchart of the method in the first embodiment.
[0034] Figure 2 This is a system structure diagram of the second embodiment.
[0035] Figure 3 The first embodiment is a schematic diagram of the global maximum power point tracking method for photovoltaic arrays based on SVM and feature mapping.
[0036] Figure 4 The flowchart shows the optimization of the support vector machine penalty optimization parameters and kernel function parameters using the particle swarm optimization algorithm in the first embodiment. Detailed Implementation
[0037] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.
[0038] It should be noted that the terminology used herein is for the purpose of describing particular implementations only and is not intended to limit the exemplary implementations of the present invention.
[0039] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.
[0040] The overall concept proposed in this invention is as follows:
[0041] The inventors discovered in their research that existing methods still cannot adequately balance hardware cost, tracking speed, tracking accuracy, and universality. Given the ability to accurately obtain the maximum power point (MPPT), reducing the hardware cost of the MPT algorithm, improving its tracking speed, and enhancing its universality have become key optimization directions for MPT technology. Currently, a fast and accurate global MPT method for photovoltaic arrays is lacking. Therefore, this invention addresses the problem that existing global maximum power point tracking (GMPT) methods still cannot effectively balance hardware cost, tracking speed, tracking accuracy, and universality. It proposes a GMPT method for photovoltaic arrays based on SVM and feature mapping. Through offline modeling, PV curves of the photovoltaic array under different shading conditions are collected, and feature point selection rules are determined based on statistical analysis. The shape characteristics of the PV curve are described using the voltage, power, and slope values corresponding to multiple feature points. A mapping relationship between the PV curve shape characteristics and the global maximum power point is established based on a support vector machine, and the global maximum power point is quickly located by scanning several feature points on the PV curve. Based on the particle swarm optimization algorithm, the optimal values of the support vector machine penalty parameter and kernel function parameter are achieved, with the prediction accuracy of the voltage at the global maximum power point as the optimization direction. Finally, high-precision GMPT is further achieved by combining an adaptive perturbation observation method.
[0042] This invention significantly improves the tracking speed of the global maximum power point of a photovoltaic array while ensuring tracking accuracy.
[0043] Example 1
[0044] This embodiment discloses a global maximum power point tracking method for photovoltaic arrays based on SVM and feature mapping.
[0045] like Figure 1 As shown, the global maximum power point tracking method for photovoltaic arrays based on SVM and feature mapping includes the following steps:
[0046] Model the photovoltaic array and collect the PV curves of the photovoltaic array under different shading conditions;
[0047] Feature points of the PV curve are selected, and the voltage, power, and slope values corresponding to the feature points are extracted.
[0048] A training set is constructed based on the extracted feature point voltage, power, and slope values. The support vector machine is then trained to obtain a well-trained support vector machine.
[0049] The parameters of the trained support vector machine are optimized to obtain the optimized support vector machine.
[0050] The output voltage and current of the photovoltaic array are sampled in real time to determine the type of shading of the photovoltaic array;
[0051] Based on the occlusion type, the maximum power point is tracked using an optimized support vector machine and an adaptive P&O algorithm.
[0052] Furthermore, the characteristic points of the PV curve are selected, specifically as follows:
[0053] Statistical analysis is used to determine the number of feature points of the PV curve. The PV curve of the photovoltaic array is divided into N+1 equal parts using N feature points, and the value of N is gradually increased from small to large until the N feature points can completely cover all the peak values of the PV curve. Then, the voltage, power and slope values corresponding to the feature points are extracted.
[0054] Furthermore, a training set is constructed based on the extracted feature point voltage, power, and slope values to train the support vector machine, resulting in a trained support vector machine, specifically:
[0055] The input parameters of the training set are the voltage, power, and slope values corresponding to the feature points on the PV curve, and the output parameter is the voltage at the maximum power point of the photovoltaic array.
[0056] Furthermore, the parameters of the trained support vector machine are optimized to obtain the optimized support vector machine, specifically as follows:
[0057] This paper uses a particle swarm optimization algorithm to optimize the values of the penalty optimization parameters and kernel function parameters of a support vector machine. The particles are set as two-dimensional parameters containing both penalty optimization parameters and kernel function parameters. Iterative optimization is performed with the prediction accuracy of the voltage at the global maximum power point as the objective function until a set number of iterations is reached. Specific steps are as follows: Figure 4 As shown.
[0058] Furthermore, the output voltage and current of the photovoltaic array are sampled in real time to determine the shading type of the photovoltaic array, specifically:
[0059] First, determine whether the photovoltaic array is being shaded. If shading has occurred, further determine whether the shading type is partial or complete based on the output voltage and output current of the photovoltaic array.
[0060] Furthermore, based on the occlusion type, and using the optimized support vector machine and adaptive P&O algorithm, global maximum power point tracking is achieved, specifically as follows:
[0061] like Figure 3As shown, if the determination result is partial occlusion, first scan the feature points and calculate the corresponding voltage, power, and slope values, input them into the optimized SVM model to obtain the rough position of the maximum power point, and then use the adaptive P&O algorithm to achieve the accurate tracking of the maximum power point;
[0062] If the determination result is complete occlusion, directly use the adaptive P&O algorithm and set the initial perturbation position to 0.8V OC , to achieve the fast tracking of the maximum power point.
[0063] Furthermore, the determination method for judging whether the photovoltaic array is occluded is: when the power change of the photovoltaic power generation system is greater than 5%, it is determined that the photovoltaic array is occluded.
[0064] Furthermore, to judge whether the occlusion type of the photovoltaic array is partial occlusion or complete occlusion, the specific determination method is exemplified as follows:
[0065] It is realized by scanning 5 evenly distributed feature points on the P-V curve. Use {dP1 / dV1, dP2 / dV2, dP3 / dV3, dP4 / dV4, dP5 / dV5} to represent the slopes of the 5 feature points, and {P1, P2, P3, P4, P5} to represent the powers corresponding to the 5 feature points. When V1 < V2 < V3 < V4 < V5, the P-V curve under the complete occlusion condition will satisfy the following characteristics:
[0066] 1) dP1 / dV1 > dP2 / dV2 > dP3 / dV3 > dP4 / dV4 > dP5 / dV5;
[0067] 2) dP / dV is positive in the first half and negative in the second half;
[0068] 3) The power P increases monotonically in the first half and decreases monotonically in the second half.
[0069] Once these three characteristics are satisfied simultaneously, it can be determined that the occlusion condition of the photovoltaic array is complete occlusion; if they cannot be satisfied simultaneously, it can be determined that the occlusion condition of the photovoltaic array is partial occlusion.
[0070] Embodiment 2
[0071] This embodiment discloses a global maximum power point tracking system for a photovoltaic array based on SVM and feature mapping.
[0072] As Figure 2 shown, the global maximum power point tracking system for a photovoltaic array based on SVM and feature mapping includes:
[0073] A P-V curve acquisition module, configured to: model the photovoltaic array and collect the P-V curves of the photovoltaic array under different occlusion conditions;
[0074] The feature point selection module is configured to: select feature points of the PV curve and extract the voltage, power and slope values corresponding to the feature points;
[0075] The training module is configured to: construct a training set based on the extracted feature point voltage, power, and slope values, train the support vector machine, and obtain a trained support vector machine;
[0076] The optimization module is configured to optimize the parameters of the trained support vector machine to obtain an optimized support vector machine.
[0077] The shading type determination module is configured to: sample the output voltage and output current of the photovoltaic array in real time and determine the shading type of the photovoltaic array;
[0078] The maximum power point tracking module is configured to track the maximum power point based on an optimized support vector machine and an adaptive P&O algorithm, depending on the occlusion type.
[0079] Example 3
[0080] The purpose of this embodiment is to provide a computer-readable storage medium.
[0081] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the photovoltaic array global maximum power point tracking method based on SVM and feature mapping as described in Embodiment 1 of this disclosure.
[0082] Example 4
[0083] The purpose of this embodiment is to provide an electronic device.
[0084] An electronic device includes a memory, a processor, and a program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps in the photovoltaic array global maximum power point tracking method based on SVM and feature mapping as described in Embodiment 1 of this disclosure.
[0085] The steps and methods involved in the apparatuses of Embodiments 2, 3, and 4 above correspond to those in Embodiment 1. For specific implementation details, please refer to the relevant description section of Embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood as including any medium capable of storing, encoding, or carrying an instruction set for execution by a processor and enabling the processor to perform any of the methods in this invention.
[0086] Those skilled in the art will understand that the modules or steps of the present invention described above can be implemented using general-purpose computer devices. Optionally, they can be implemented using computer-executable program code, thereby allowing them to be stored in a storage device for execution by a computer device, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. The present invention is not limited to any particular combination of hardware and software.
[0087] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.
Claims
1. A photovoltaic array global maximum power point tracking method based on SVM and feature mapping, characterized in that, Includes the following steps: Model the photovoltaic array and collect the PV curves of the photovoltaic array under different shading conditions; Feature points of the PV curve are selected, and the voltage, power, and slope values corresponding to the feature points are extracted. A training set is constructed based on the extracted feature point voltage, power, and slope values. The support vector machine is then trained to obtain a well-trained support vector machine. Specifically: The input parameters of the training set are the voltage, power, and slope values corresponding to the feature points on the PV curve, and the output parameter is the voltage at the maximum power point of the photovoltaic array. The parameters of the trained support vector machine are optimized to obtain the optimized support vector machine. The output voltage and current of the photovoltaic array are sampled in real time to determine the type of shading of the photovoltaic array; Based on the occlusion type, the maximum power point is tracked using an optimized support vector machine and an adaptive P&O algorithm. If the determination result is partial occlusion, first scan the feature points and calculate the corresponding voltage, power and slope values, input them into the optimized SVM model to obtain the approximate location of the maximum power point, and then use the adaptive P&O algorithm to achieve accurate tracking of the maximum power point.
2. The photovoltaic array global maximum power point tracking method based on SVM and feature mapping as described in claim 1, characterized in that, The characteristic points of the PV curve are selected as follows: Statistical analysis is used to determine the number of feature points of the PV curve. The PV curve of the photovoltaic array is divided into N+1 equal parts using N feature points, and the value of N is gradually increased from small to large until the N feature points can completely cover all the peak values of the PV curve. Then, the voltage, power and slope values corresponding to the feature points are extracted.
3. The photovoltaic array global maximum power point tracking method based on SVM and feature mapping as described in claim 1, characterized in that, The parameters of the trained support vector machine are optimized to obtain the optimized support vector machine, specifically: The particle swarm optimization algorithm is used to optimize the values of the penalty optimization parameters and kernel function parameters of the support vector machine. The particles are set as two-dimensional parameters containing the penalty optimization parameters and kernel function parameters. The prediction accuracy of the voltage at the global maximum power point is used as the objective function for iterative optimization until the set number of iterations is reached.
4. The photovoltaic array global maximum power point tracking method based on SVM and feature mapping as described in claim 1, characterized in that, The output voltage and current of the photovoltaic array are sampled in real time to determine the shading type of the photovoltaic array. Specifically: First, determine whether the photovoltaic array is being shaded. If shading has occurred, further determine whether the shading type is partial or complete based on the output voltage and output current of the photovoltaic array.
5. The photovoltaic array global maximum power point tracking method based on SVM and feature mapping as described in claim 4, characterized in that, Based on the occlusion type, and using an optimized support vector machine and adaptive P&O algorithm, global maximum power point tracking is achieved, specifically as follows: If the determination result is complete occlusion, then the adaptive P&O algorithm is directly used, and the initial perturbation position is set to 0.8V. OC This enables rapid tracking of the maximum power point.
6. The photovoltaic array global maximum power point tracking method based on SVM and feature mapping as described in claim 4, characterized in that, The method for determining whether a photovoltaic array has been shaded is as follows: when the power change of the photovoltaic power generation system is greater than 5%, the photovoltaic array is considered to have been shaded.
7. A photovoltaic array global maximum power point tracking system based on SVM and feature mapping, characterized in that: include: The PV curve acquisition module is configured to: model the photovoltaic array and collect the PV curves of the photovoltaic array under different shading conditions; The feature point selection module is configured to: select feature points of the PV curve and extract the voltage, power and slope values corresponding to the feature points; The training module is configured to: construct a training set based on the extracted feature point voltage, power, and slope values; train the support vector machine to obtain a trained support vector machine; specifically: The input parameters of the training set are the voltage, power, and slope values corresponding to the feature points on the PV curve, and the output parameter is the voltage at the maximum power point of the photovoltaic array. The optimization module is configured to optimize the parameters of the trained support vector machine to obtain an optimized support vector machine. The shading type determination module is configured to: sample the output voltage and output current of the photovoltaic array in real time and determine the shading type of the photovoltaic array; The maximum power point tracking module is configured to track the maximum power point based on an optimized support vector machine and an adaptive P&O algorithm, according to the occlusion type. If the determination result is partial occlusion, first scan the feature points and calculate the corresponding voltage, power and slope values, input them into the optimized SVM model to obtain the approximate location of the maximum power point, and then use the adaptive P&O algorithm to achieve accurate tracking of the maximum power point.
8. A computer-readable storage medium having a program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps in the photovoltaic array global maximum power point tracking method based on SVM and feature mapping as described in any one of claims 1-6.
9. An electronic device, comprising a memory, a processor, and a program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps in the photovoltaic array global maximum power point tracking method based on SVM and feature mapping as described in any one of claims 1-6.