Photovoltaic module fault diagnosis method and system based on IV-CV feature collaborative fusion

Abstract

This invention relates to photovoltaic (PV) module fault diagnosis, specifically to a PV module fault diagnosis method and system based on IV-CV feature synergistic fusion. The method involves: acquiring the voltage and current of multiple PV modules in a PV array and identifying abnormal PV modules through comparison; acquiring the I-V curves of the abnormal PV cells and inputting them into a fault diagnosis model to obtain a first fault diagnosis result; acquiring the original image and infrared image of the abnormal PV module, and performing recognition processing on the original image based on the infrared image to obtain a fault area image; using a diagnostic matching matrix to perform fault diagnosis on the fault area image to obtain a second fault diagnosis result; combining the first and second fault diagnosis results to obtain the final fault diagnosis result of the abnormal PV module; and adaptively recommending targeted fault repair solutions based on the final fault diagnosis result of the abnormal PV module. This invention overcomes the shortcomings of accurately and comprehensively diagnosing PV module faults.

Description

Technical Field

[0001] This invention relates to photovoltaic module fault diagnosis, specifically to a photovoltaic module fault diagnosis method and system based on IV-CV feature synergistic fusion. Background Technology

[0002] With the continuous growth of global demand for clean energy, photovoltaic power generation, as an important way to utilize renewable energy, has experienced rapid development and widespread application. As the core component of a photovoltaic power generation system, the operating status of photovoltaic modules directly affects the power generation efficiency and stability of the entire system.

[0003] During the long-term operation of photovoltaic modules, various faults inevitably occur due to environmental factors (such as high temperature, high humidity, and ultraviolet radiation), electrical stress, and aging. These faults include hot spots, microcracks, and PID effects. These faults can lead to a decrease in the output power of photovoltaic modules and may even cause safety problems.

[0004] Currently, fault diagnosis methods for photovoltaic (PV) modules are mainly divided into two categories: electrical characteristic analysis and image recognition analysis. Electrical characteristic analysis is usually based on the PV module's IV curve, analyzing the curve's characteristic parameters to determine the presence of faults. However, this method struggles to accurately identify some relatively complex fault types. Image recognition analysis utilizes visible light and infrared images to detect surface defects and abnormal heating areas in PV modules. However, single image analysis can be affected by factors such as image quality and environmental interference, leading to inaccurate diagnostic results.

[0005] Therefore, in order to diagnose photovoltaic module faults more accurately and comprehensively, and improve the power generation efficiency and stability of photovoltaic power generation systems, it is necessary to study a photovoltaic module fault diagnosis method and system based on the synergistic fusion of IV-CV features. Summary of the Invention

[0006] (a) Technical problems to be solved

[0007] In view of the above-mentioned shortcomings of the existing technology, the present invention provides a photovoltaic module fault diagnosis method and system based on IV-CV feature synergistic fusion, which can effectively overcome the defects of the existing technology in that it is difficult to accurately and comprehensively diagnose photovoltaic module faults.

[0008] (II) Technical Solution

[0009] To achieve the above objectives, the present invention provides the following technical solution:

[0010] A photovoltaic module fault diagnosis method based on IV-CV feature synergy fusion includes the following steps:

[0011] S1. Collect the voltage and current of multiple photovoltaic modules in the photovoltaic array, and identify abnormal photovoltaic modules by comparison;

[0012] S2. Collect the IV curve of the abnormal photovoltaic cell and input it into the fault diagnosis model to obtain the first fault diagnosis result;

[0013] S3. Collect raw images and infrared images of abnormal photovoltaic modules, and perform identification processing on the raw images based on the infrared images to obtain images of the fault areas;

[0014] S4. Use the diagnostic matching matrix to perform fault diagnosis on the fault area image to obtain the second fault diagnosis result;

[0015] S5. Combining the first fault diagnosis result and the second fault diagnosis result, the final fault diagnosis result of the abnormal photovoltaic module is obtained;

[0016] S6. Based on the final fault diagnosis results of the abnormal photovoltaic modules, adaptively recommend targeted fault repair solutions.

[0017] Preferably, S1 involves collecting the voltage and current of multiple photovoltaic modules in the photovoltaic array, including:

[0018] The photovoltaic array includes multiple parallel branches connected in series, each parallel branch includes at least two photovoltaic modules connected in parallel, each parallel branch is connected in parallel with a voltage sensor, and each photovoltaic module is connected in series with a current sensor.

[0019] Voltage sensors are used to collect the voltage of all photovoltaic modules in the parallel branch, and current sensors are used to collect the current of all photovoltaic modules.

[0020] Preferably, in S1, identifying abnormal photovoltaic modules through comparison includes:

[0021] By comparing the voltages collected by voltage sensors in multiple parallel branches, the target parallel branch where the abnormal photovoltaic module is located can be determined.

[0022] Abnormal photovoltaic modules are identified by comparing the currents collected by all current sensors in the target parallel branch.

[0023] Preferably, the IV curve of the abnormal photovoltaic cell is collected in S2, including:

[0024] For each photovoltaic cell in the abnormal photovoltaic module, all other photovoltaic cells in the abnormal photovoltaic module are stopped from operating, and the open-circuit voltage and short-circuit current of each photovoltaic cell are measured.

[0025] Abnormal photovoltaic cells are identified by comparing the open-circuit voltage and short-circuit current of all photovoltaic cells.

[0026] Based on a single diode model, the IV curve of an abnormal photovoltaic cell is generated according to its open-circuit voltage and short-circuit current.

[0027] Preferably, the fault diagnosis model is input into S2 to obtain the first fault diagnosis result, including:

[0028] Collect the current IV curve, temperature and illuminance of the abnormal photovoltaic cell, and obtain the feature vector of the abnormal photovoltaic cell after preprocessing.

[0029] The feature vectors of the abnormal photovoltaic cells are input into a pre-trained fault diagnosis model to obtain the first fault diagnosis result.

[0030] Preferably, before obtaining the first fault diagnosis result by inputting the fault diagnosis model in S2, the following steps are included:

[0031] S21. Divide the historical dataset into training set, validation set and test set according to the preset ratio;

[0032] S22. Define the loss function and optimizer for the fault diagnosis model;

[0033] S23. Input the training set into the fault diagnosis model for model training;

[0034] S24. Calculate the loss value based on the loss function, and the optimizer updates the model parameters based on the loss value and network gradient information;

[0035] S25. If the loss value is less than the preset threshold, the model training ends and the current fault diagnosis model is the trained fault diagnosis model. Otherwise, return to S23 and continue to train the model using the training set.

[0036] S26. Input the validation set into the trained fault diagnosis model, evaluate the model's generalization ability by observing its performance on the validation set, and fine-tune the model's hyperparameters and structure.

[0037] S27. Input the test set into the optimized fault diagnosis model, evaluate the model performance, and finally obtain the pre-trained fault diagnosis model.

[0038] The fault diagnosis model is built based on a multilayer perceptron (MLP).

[0039] Preferably, in step S3, the original image and infrared image of the abnormal photovoltaic module are acquired, and the original image is processed based on the infrared image to obtain a fault area image, including:

[0040] The original image and infrared image of the abnormal photovoltaic module are collected, and the infrared image is pixelated to obtain pixel data.

[0041] An overlay fitting operation is performed on the pixel data and the original image to obtain the image of the fault area.

[0042] Preferably, in S4, a diagnostic matching matrix is ​​used to perform fault diagnosis on the fault area image to obtain a second fault diagnosis result, including:

[0043] The diagnostic matching matrix D is represented by the following formula:

[0044] ;

[0045] Among them, I1~I m For m types of fault area images, F1~F n For n possible fault diagnosis results, an element in the diagnosis matching matrix D is 1 if the fault area image has a corresponding fault diagnosis result, and 0 if the fault area image does not have a corresponding fault diagnosis result.

[0046] The fault diagnosis result is obtained by using the diagnostic matching matrix D to perform fault diagnosis on the fault area image.

[0047] Preferably, in step S6, based on the final fault diagnosis result of the abnormal photovoltaic module, an adaptive and targeted fault repair plan is recommended, including:

[0048] Based on the final fault diagnosis results of the abnormal photovoltaic modules, the repair entity is determined, wherein the repair entity includes at least one of repair personnel, repair robots and repair drones;

[0049] By combining the final fault diagnosis results of the abnormal photovoltaic modules with the repair entity, the corresponding fault repair plan is matched from the repair information database.

[0050] The photovoltaic module fault diagnosis system based on IV-CV feature synergy fusion includes a first data acquisition module, an abnormal photovoltaic module identification module, an abnormal photovoltaic cell identification module, a second data acquisition module, a first fault diagnosis result generation module, an image acquisition module, a fault area image generation module, a second fault diagnosis result generation module, a final fault diagnosis result generation module, and a fault repair plan recommendation module.

[0051] The first data acquisition module acquires the voltage and current of multiple photovoltaic modules in the photovoltaic array. The photovoltaic array includes multiple parallel branches connected in series. Each parallel branch includes at least two photovoltaic modules connected in parallel. Each parallel branch is connected in parallel with a voltage sensor, and each photovoltaic module is connected in series with a current sensor. The voltage sensor is used to acquire the voltage of all photovoltaic modules in the parallel branch, and the current sensor is used to acquire the current of all photovoltaic modules.

[0052] The abnormal photovoltaic module identification module determines the target parallel branch where the abnormal photovoltaic module is located by comparing the voltage collected by voltage sensors of multiple parallel branches, and determines the abnormal photovoltaic module by comparing the current collected by all current sensors in the target parallel branch.

[0053] The abnormal photovoltaic cell identification module stops all other photovoltaic cells in the abnormal photovoltaic module for each photovoltaic cell in the abnormal photovoltaic module, measures the open-circuit voltage and short-circuit current of each photovoltaic cell, and identifies the abnormal photovoltaic cell by comparing the open-circuit voltage and short-circuit current of all photovoltaic cells.

[0054] The second data acquisition module, based on a single diode model, generates the IV curve of the abnormal photovoltaic cell according to the open-circuit voltage and short-circuit current of the abnormal photovoltaic cell.

[0055] The first fault diagnosis result generation module collects the current IV curve, temperature and illuminance of the abnormal photovoltaic cell, and obtains the feature vector of the abnormal photovoltaic cell after preprocessing. The feature vector of the abnormal photovoltaic cell is then input into the pre-trained fault diagnosis model to obtain the first fault diagnosis result.

[0056] The image acquisition module acquires raw and infrared images of the abnormal photovoltaic modules;

[0057] The fault area image generation module performs pixelation processing on the infrared image to obtain pixel data, and performs an overlap fitting operation on the pixel data and the original image to obtain the fault area image.

[0058] The second fault diagnosis result generation module uses a diagnosis matching matrix to perform fault diagnosis on the fault area image and obtains the second fault diagnosis result.

[0059] The final fault diagnosis result generation module combines the first fault diagnosis result and the second fault diagnosis result to obtain the final fault diagnosis result of the abnormal photovoltaic module.

[0060] The fault repair solution recommendation module determines the repair entity based on the final fault diagnosis results of the abnormal photovoltaic module, and matches the corresponding fault repair solution from the repair information database by combining the final fault diagnosis results of the abnormal photovoltaic module and the repair entity.

[0061] (III) Beneficial Effects

[0062] Compared with existing technologies, the photovoltaic module fault diagnosis method and system based on IV-CV feature synergistic fusion provided by this invention has the following beneficial effects:

[0063] 1) Integrating multiple features to improve the accuracy of fault diagnosis

[0064] By synergistically fusing IV-CV features for photovoltaic module fault diagnosis, the advantages of both electrical characteristics and image features are fully utilized. On the one hand, the IV curves of abnormal photovoltaic cells are acquired and analyzed using a fault diagnosis model built on a multilayer perceptron (MLP), which can deeply explore fault information from an electrical perspective. On the other hand, the original images and infrared images of abnormal photovoltaic modules are acquired, and after processing to obtain images of the fault area, fault diagnosis is performed using a diagnostic matching matrix, which can provide fault clues at the image level. This multi-dimensional information fusion diagnosis method effectively overcomes the limitations of single diagnostic methods, greatly improves the accuracy and reliability of fault diagnosis, and can more accurately identify the fault type and location of photovoltaic modules.

[0065] 2) Precisely match solutions to improve fault repair efficiency.

[0066] Based on the final fault diagnosis results of abnormal photovoltaic modules, the system can adaptively recommend targeted fault repair solutions. By identifying appropriate repair entities (such as repair personnel, repair robots, or repair drones) and accurately matching the corresponding fault repair solutions from the repair information database in conjunction with the final fault diagnosis results, the system avoids the behavior of repair personnel blindly troubleshooting and attempting repairs. This not only saves repair time and reduces power generation losses caused by faults, but also enables repair resources to be used more rationally, significantly improving the fault repair efficiency of photovoltaic modules.

[0067] 3) Intelligent automatic processing enhances the system's intelligence level.

[0068] The entire technical solution, from data acquisition and fault diagnosis to solution recommendation, achieves a high degree of intelligence and automation: In the data acquisition stage, various sensors are used to automatically collect information such as voltage, current, and images of photovoltaic modules; during the fault diagnosis process, the fault diagnosis model and diagnostic matching matrix can automatically analyze and process the input data to obtain fault diagnosis results; in the solution recommendation stage, the system can automatically determine the repair entity and match the corresponding fault repair solution based on the final fault diagnosis result. This intelligent and automated processing flow significantly reduces manual intervention, reduces the impact of human factors on fault diagnosis and repair, and effectively enhances the intelligence level of photovoltaic module fault diagnosis and repair. Attached Figure Description

[0069] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0070] Figure 1This is a schematic diagram of the process of the present invention;

[0071] Figure 2 This is a schematic diagram of the process for obtaining the first fault diagnosis result in this invention;

[0072] Figure 3 This is a schematic diagram of the process for obtaining the second fault diagnosis result in this invention. Detailed Implementation

[0073] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0074] The following describes the specific process of the photovoltaic module fault diagnosis method based on IV-CV feature synergistic fusion provided by this invention, using a specific example (e.g.) Figure 1 (as shown) and technical effects.

[0075] S1. Collect the voltage and current of multiple photovoltaic modules in the photovoltaic array, and identify abnormal photovoltaic modules by comparison.

[0076] 1) Collect the voltage and current of multiple photovoltaic modules in the photovoltaic array, such as... Figure 2 As shown, it includes:

[0077] The photovoltaic array includes multiple parallel branches connected in series, each parallel branch includes at least two photovoltaic modules connected in parallel, each parallel branch is connected in parallel with a voltage sensor, and each photovoltaic module is connected in series with a current sensor.

[0078] Voltage sensors are used to collect the voltage of all photovoltaic modules in the parallel branch, and current sensors are used to collect the current of all photovoltaic modules.

[0079] 2) Identify abnormal photovoltaic modules through comparison, such as Figure 2 As shown, it includes:

[0080] By comparing the voltages collected by voltage sensors in multiple parallel branches, the target parallel branch where the abnormal photovoltaic module is located can be determined.

[0081] Abnormal photovoltaic modules are identified by comparing the currents collected by all current sensors in the target parallel branch.

[0082] S2. Collect the IV curve of the abnormal photovoltaic cell and input it into the fault diagnosis model to obtain the first fault diagnosis result.

[0083] 1) Collect the IV curves of abnormal photovoltaic cells, such as... Figure 2 As shown, it includes:

[0084] For each photovoltaic cell in the abnormal photovoltaic module, all other photovoltaic cells in the abnormal photovoltaic module are stopped from operating, and the open-circuit voltage and short-circuit current of each photovoltaic cell are measured.

[0085] Abnormal photovoltaic cells are identified by comparing the open-circuit voltage and short-circuit current of all photovoltaic cells.

[0086] Based on a single diode model, the IV curve of an abnormal photovoltaic cell is generated according to its open-circuit voltage and short-circuit current.

[0087] 2) Input the fault diagnosis model to obtain the first fault diagnosis result, such as... Figure 2 As shown, it includes:

[0088] Collect the current IV curve, temperature and illuminance of the abnormal photovoltaic cell, and obtain the feature vector of the abnormal photovoltaic cell after preprocessing.

[0089] The feature vectors of the abnormal photovoltaic cells are input into a pre-trained fault diagnosis model to obtain the first fault diagnosis result.

[0090] Specifically, before obtaining the first fault diagnosis result from the input fault diagnosis model, the following steps are included:

[0091] S21. Divide the historical dataset into training set, validation set and test set according to the preset ratio;

[0092] S22. Define the loss function and optimizer for the fault diagnosis model;

[0093] S23. Input the training set into the fault diagnosis model for model training;

[0094] S24. Calculate the loss value based on the loss function, and the optimizer updates the model parameters based on the loss value and network gradient information;

[0095] S25. If the loss value is less than the preset threshold, the model training ends and the current fault diagnosis model is the trained fault diagnosis model. Otherwise, return to S23 and continue to train the model using the training set.

[0096] S26. Input the validation set into the trained fault diagnosis model, evaluate the model's generalization ability by observing its performance on the validation set, and fine-tune the model's hyperparameters and structure.

[0097] S27. Input the test set into the optimized fault diagnosis model, evaluate the model performance, and finally obtain the pre-trained fault diagnosis model.

[0098] The fault diagnosis model is built based on a multilayer perceptron (MLP).

[0099] S3. Acquire raw and infrared images of the abnormal photovoltaic modules, and perform identification processing on the raw images based on the infrared images to obtain images of the fault areas, such as... Figure 3 As shown, it includes:

[0100] The original image and infrared image of the abnormal photovoltaic module are collected, and the infrared image is pixelated to obtain pixel data.

[0101] An overlay fitting operation is performed on the pixel data and the original image to obtain the image of the fault area.

[0102] S4. Use the diagnostic matching matrix to perform fault diagnosis on the fault area image to obtain the second fault diagnosis result, such as... Figure 3 As shown, it includes:

[0103] The diagnostic matching matrix D is represented by the following formula:

[0104] ;

[0105] Among them, I1~I m For m types of fault area images, F1~F n For n possible fault diagnosis results, an element in the diagnosis matching matrix D is 1 if the fault area image has a corresponding fault diagnosis result, and 0 if the fault area image does not have a corresponding fault diagnosis result.

[0106] The fault diagnosis result is obtained by using the diagnostic matching matrix D to perform fault diagnosis on the fault area image.

[0107] The technical solution of this application uses the synergistic fusion of IV-CV features for photovoltaic module fault diagnosis, making full use of the advantages of electrical characteristics and image features. On the one hand, it acquires the IV curves of abnormal photovoltaic cells and analyzes them using a fault diagnosis model built based on a multilayer perceptron (MLP), which can deeply mine fault information from an electrical perspective. On the other hand, it acquires the original image and infrared image of the abnormal photovoltaic module, processes them to obtain the fault area image, and then uses a diagnostic matching matrix for fault diagnosis, which can provide fault clues at the image level. This multi-dimensional information fusion diagnosis method effectively overcomes the limitations of single diagnosis methods, greatly improves the accuracy and reliability of fault diagnosis, and can more accurately identify the fault type and location of photovoltaic modules.

[0108] S5. Combining the first fault diagnosis result and the second fault diagnosis result, the final fault diagnosis result of the abnormal photovoltaic module is obtained.

[0109] S6. Based on the final fault diagnosis results of the abnormal photovoltaic modules, adaptively recommend targeted fault repair solutions, including:

[0110] Based on the final fault diagnosis results of the abnormal photovoltaic modules, the repair entity is determined, wherein the repair entity includes at least one of repair personnel, repair robots and repair drones;

[0111] By combining the final fault diagnosis results of the abnormal photovoltaic modules with the repair entity, the corresponding fault repair plan is matched from the repair information database.

[0112] The above technical solution can adaptively recommend targeted fault repair solutions based on the final fault diagnosis results of abnormal photovoltaic modules. By identifying suitable repair entities (such as repair personnel, repair robots, or repair drones) and accurately matching the corresponding fault repair solutions from the repair information database in conjunction with the final fault diagnosis results, it avoids the behavior of repair personnel blindly troubleshooting and attempting repairs. This not only saves repair time and reduces power generation losses caused by faults, but also enables more rational use of repair resources and significantly improves the fault repair efficiency of photovoltaic modules.

[0113] Based on the aforementioned photovoltaic module fault diagnosis method based on IV-CV feature synergy fusion, this application also discloses a photovoltaic module fault diagnosis system based on IV-CV feature synergy fusion, including a first data acquisition module, an abnormal photovoltaic module determination module, an abnormal photovoltaic cell determination module, a second data acquisition module, a first fault diagnosis result generation module, an image acquisition module, a fault area image generation module, a second fault diagnosis result generation module, a final fault diagnosis result generation module, and a fault repair plan recommendation module;

[0114] The first data acquisition module acquires the voltage and current of multiple photovoltaic modules in the photovoltaic array. The photovoltaic array includes multiple parallel branches connected in series. Each parallel branch includes at least two photovoltaic modules connected in parallel. Each parallel branch is connected in parallel with a voltage sensor, and each photovoltaic module is connected in series with a current sensor. The voltage sensor is used to acquire the voltage of all photovoltaic modules in the parallel branch, and the current sensor is used to acquire the current of all photovoltaic modules.

[0115] The abnormal photovoltaic module identification module determines the target parallel branch where the abnormal photovoltaic module is located by comparing the voltage collected by voltage sensors of multiple parallel branches, and determines the abnormal photovoltaic module by comparing the current collected by all current sensors in the target parallel branch.

[0116] The abnormal photovoltaic cell identification module stops all other photovoltaic cells in the abnormal photovoltaic module for each photovoltaic cell in the abnormal photovoltaic module, measures the open-circuit voltage and short-circuit current of each photovoltaic cell, and identifies the abnormal photovoltaic cell by comparing the open-circuit voltage and short-circuit current of all photovoltaic cells.

[0117] The second data acquisition module, based on a single diode model, generates the IV curve of the abnormal photovoltaic cell according to the open-circuit voltage and short-circuit current of the abnormal photovoltaic cell.

[0118] The first fault diagnosis result generation module collects the current IV curve, temperature and illuminance of the abnormal photovoltaic cell, and obtains the feature vector of the abnormal photovoltaic cell after preprocessing. The feature vector of the abnormal photovoltaic cell is then input into the pre-trained fault diagnosis model to obtain the first fault diagnosis result.

[0119] The image acquisition module acquires raw and infrared images of the abnormal photovoltaic modules;

[0120] The fault area image generation module performs pixelation processing on the infrared image to obtain pixel data, and performs an overlap fitting operation on the pixel data and the original image to obtain the fault area image.

[0121] The second fault diagnosis result generation module uses a diagnosis matching matrix to perform fault diagnosis on the fault area image and obtains the second fault diagnosis result.

[0122] The final fault diagnosis result generation module combines the first fault diagnosis result and the second fault diagnosis result to obtain the final fault diagnosis result of the abnormal photovoltaic module.

[0123] The fault repair solution recommendation module determines the repair entity based on the final fault diagnosis results of the abnormal photovoltaic module, and matches the corresponding fault repair solution from the repair information database by combining the final fault diagnosis results of the abnormal photovoltaic module and the repair entity.

[0124] The technical solution of this application achieves a high degree of intelligence and automation from data acquisition and fault diagnosis to solution recommendation: In the data acquisition stage, various sensors are used to automatically collect information such as voltage, current, and images of photovoltaic modules; in the fault diagnosis process, the fault diagnosis model and diagnosis matching matrix can automatically analyze and process the input data to obtain fault diagnosis results; in the solution recommendation stage, the system can automatically determine the repair entity and match the corresponding fault repair solution based on the final fault diagnosis result. This intelligent and automated processing flow significantly reduces manual intervention, reduces the impact of human factors on fault diagnosis and repair, and effectively enhances the intelligence level of photovoltaic module fault diagnosis and repair.

[0125] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions will 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 invention.

Claims

1. A photovoltaic module fault diagnosis method based on IV-CV feature synergistic fusion, characterized in that: Includes the following steps: S1. Collect the voltage and current of multiple photovoltaic modules in the photovoltaic array, and identify abnormal photovoltaic modules by comparison; S2. Collect the IV curve of the abnormal photovoltaic cell and input it into the fault diagnosis model to obtain the first fault diagnosis result; S3. Collect raw images and infrared images of abnormal photovoltaic modules, and perform identification processing on the raw images based on the infrared images to obtain images of the fault areas; S4. Use the diagnostic matching matrix to perform fault diagnosis on the fault area image to obtain the second fault diagnosis result; S5. Combining the first fault diagnosis result and the second fault diagnosis result, the final fault diagnosis result of the abnormal photovoltaic module is obtained; S6. Based on the final fault diagnosis results of the abnormal photovoltaic modules, adaptively recommend targeted fault repair solutions.

2. The photovoltaic module fault diagnosis method based on IV-CV feature synergistic fusion according to claim 1, characterized in that: S1 collects the voltage and current of multiple photovoltaic modules in the photovoltaic array, including: The photovoltaic array includes multiple parallel branches connected in series, each parallel branch includes at least two photovoltaic modules connected in parallel, each parallel branch is connected in parallel with a voltage sensor, and each photovoltaic module is connected in series with a current sensor. Voltage sensors are used to collect the voltage of all photovoltaic modules in the parallel branch, and current sensors are used to collect the current of all photovoltaic modules.

3. The photovoltaic module fault diagnosis method based on IV-CV feature synergistic fusion according to claim 2, characterized in that: S1 identifies abnormal photovoltaic modules through comparison, including: By comparing the voltages collected by voltage sensors in multiple parallel branches, the target parallel branch where the abnormal photovoltaic module is located can be determined. Abnormal photovoltaic modules are identified by comparing the currents collected by all current sensors in the target parallel branch.

4. The photovoltaic module fault diagnosis method based on IV-CV feature synergistic fusion according to claim 3, characterized in that: The IV curves of abnormal photovoltaic cells collected in S2 include: For each photovoltaic cell in the abnormal photovoltaic module, all other photovoltaic cells in the abnormal photovoltaic module are stopped from operating, and the open-circuit voltage and short-circuit current of each photovoltaic cell are measured. Abnormal photovoltaic cells are identified by comparing the open-circuit voltage and short-circuit current of all photovoltaic cells. Based on a single diode model, the IV curve of an abnormal photovoltaic cell is generated according to its open-circuit voltage and short-circuit current.

5. The photovoltaic module fault diagnosis method based on IV-CV feature synergistic fusion according to claim 4, characterized in that: Input the fault diagnosis model into S2 to obtain the first fault diagnosis result, including: Collect the current IV curve, temperature and illuminance of the abnormal photovoltaic cell, and obtain the feature vector of the abnormal photovoltaic cell after preprocessing. The feature vectors of the abnormal photovoltaic cells are input into a pre-trained fault diagnosis model to obtain the first fault diagnosis result.

6. The photovoltaic module fault diagnosis method based on IV-CV feature synergistic fusion according to claim 5, characterized in that: Before obtaining the first fault diagnosis result from the input fault diagnosis model in S2, the following steps are included: S21. Divide the historical dataset into training set, validation set and test set according to the preset ratio; S22. Define the loss function and optimizer for the fault diagnosis model; S23. Input the training set into the fault diagnosis model for model training; S24. Calculate the loss value based on the loss function, and the optimizer updates the model parameters based on the loss value and network gradient information; S25. If the loss value is less than the preset threshold, the model training ends and the current fault diagnosis model is the trained fault diagnosis model. Otherwise, return to S23 and continue to train the model using the training set. S26. Input the validation set into the trained fault diagnosis model, evaluate the model's generalization ability by observing its performance on the validation set, and fine-tune the model's hyperparameters and structure. S27. Input the test set into the optimized fault diagnosis model, evaluate the model performance, and finally obtain the pre-trained fault diagnosis model. The fault diagnosis model is built based on a multilayer perceptron (MLP).

7. The photovoltaic module fault diagnosis method based on IV-CV feature synergistic fusion according to claim 1, characterized in that: S3 acquires raw and infrared images of the abnormal photovoltaic modules, and performs identification processing on the raw images based on the infrared images to obtain images of the fault areas, including: The original image and infrared image of the abnormal photovoltaic module are collected, and the infrared image is pixelated to obtain pixel data. An overlay fitting operation is performed on the pixel data and the original image to obtain the image of the fault area.

8. The photovoltaic module fault diagnosis method based on IV-CV feature synergistic fusion according to claim 7, characterized in that: In S4, a diagnostic matching matrix is ​​used to diagnose the fault area image, resulting in a second fault diagnosis result, including: The diagnostic matching matrix D is represented by the following formula: ； Among them, I1~I m For m types of fault area images, F1~F n For n possible fault diagnosis results, an element in the diagnosis matching matrix D is 1 if the fault area image has a corresponding fault diagnosis result, and 0 if the fault area image does not have a corresponding fault diagnosis result. The fault diagnosis result is obtained by using the diagnostic matching matrix D to perform fault diagnosis on the fault area image.

9. The photovoltaic module fault diagnosis method based on IV-CV feature synergistic fusion according to claim 1, characterized in that: Based on the final fault diagnosis results of the abnormal photovoltaic modules, S6 adaptively recommends targeted fault repair solutions, including: Based on the final fault diagnosis results of the abnormal photovoltaic modules, the repair entity is determined, wherein the repair entity includes at least one of repair personnel, repair robots and repair drones; By combining the final fault diagnosis results of the abnormal photovoltaic modules with the repair entity, the corresponding fault repair plan is matched from the repair information database.

10. A photovoltaic module fault diagnosis system based on IV-CV feature synergistic fusion, used to execute the photovoltaic module fault diagnosis method based on IV-CV feature synergistic fusion as described in claim 1, characterized in that: It includes a first data acquisition module, an abnormal photovoltaic module identification module, an abnormal photovoltaic cell identification module, a second data acquisition module, a first fault diagnosis result generation module, an image acquisition module, a fault area image generation module, a second fault diagnosis result generation module, a final fault diagnosis result generation module, and a fault repair plan recommendation module; The first data acquisition module acquires the voltage and current of multiple photovoltaic modules in the photovoltaic array. The photovoltaic array includes multiple parallel branches connected in series. Each parallel branch includes at least two photovoltaic modules connected in parallel. Each parallel branch is connected in parallel with a voltage sensor, and each photovoltaic module is connected in series with a current sensor. The voltage sensor is used to acquire the voltage of all photovoltaic modules in the parallel branch, and the current sensor is used to acquire the current of all photovoltaic modules. The abnormal photovoltaic module identification module determines the target parallel branch where the abnormal photovoltaic module is located by comparing the voltage collected by voltage sensors of multiple parallel branches, and determines the abnormal photovoltaic module by comparing the current collected by all current sensors in the target parallel branch. The abnormal photovoltaic cell identification module stops all other photovoltaic cells in the abnormal photovoltaic module for each photovoltaic cell in the abnormal photovoltaic module, measures the open-circuit voltage and short-circuit current of each photovoltaic cell, and identifies the abnormal photovoltaic cell by comparing the open-circuit voltage and short-circuit current of all photovoltaic cells. The second data acquisition module, based on a single diode model, generates the IV curve of the abnormal photovoltaic cell according to the open-circuit voltage and short-circuit current of the abnormal photovoltaic cell. The first fault diagnosis result generation module collects the current IV curve, temperature and illuminance of the abnormal photovoltaic cell, and obtains the feature vector of the abnormal photovoltaic cell after preprocessing. The feature vector of the abnormal photovoltaic cell is then input into the pre-trained fault diagnosis model to obtain the first fault diagnosis result. The image acquisition module acquires raw and infrared images of the abnormal photovoltaic modules; The fault area image generation module performs pixelation processing on the infrared image to obtain pixel data, and performs an overlap fitting operation on the pixel data and the original image to obtain the fault area image. The second fault diagnosis result generation module uses a diagnosis matching matrix to perform fault diagnosis on the fault area image and obtains the second fault diagnosis result. The final fault diagnosis result generation module combines the first fault diagnosis result and the second fault diagnosis result to obtain the final fault diagnosis result of the abnormal photovoltaic module. The fault repair solution recommendation module determines the repair entity based on the final fault diagnosis results of the abnormal photovoltaic module, and matches the corresponding fault repair solution from the repair information database by combining the final fault diagnosis results of the abnormal photovoltaic module and the repair entity.

Images