Photovoltaic inverter fault early warning method, device, equipment and medium

Abstract

The invention discloses a photovoltaic inverter fault early warning method, device, equipment and medium, and relates to the technical field of photovoltaic power generation, and the method comprises the following steps: carrying out the correlation analysis and correlation test simulation of various physical variables and different fault types during the operation of a photovoltaic inverter, and determining a plurality of candidate physical variables based on the simulation result; collecting fault data and normal data corresponding to the plurality of candidate physical variables, and constructing a database based on the fault data and the normal data; training the convolutional neural network model through a database to obtain a fault early warning model; and acquiring a plurality of candidate physical variable data of the photovoltaic inverter during operation in real time, and inputting the candidate physical variable data into the fault early warning model to obtain a fault early warning result. According to the method, only the data of the candidate physical variables need to be collected to train the fault early warning model, the scale of data collection is greatly reduced, the data quality is improved, meanwhile, the nonlinear fitting capacity of the data is improved, and the accuracy of an early warning result is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic power generation, and in particular to a photovoltaic inverter fault early warning method, device, equipment and medium. Background Art

[0002] The inverter is a key power conversion device in the photovoltaic power generation system. It has multiple functions such as power conversion, power quality control, system safety protection and maximum power point tracking. It plays a vital role in the performance, safety and reliability of the entire photovoltaic power generation system. The early warning and detection of traditional photovoltaic inverter faults mainly rely on safety monitoring and manual inspections during the equipment installation process. During the deployment and installation phase of the system, standard and safe and reliable power electronic devices are used to ensure the safety of the system, and then the standardized assembly process is used to achieve reliable collaborative operation between different devices. During the system operation phase, potential faults caused by complex and changeable uncertain factors such as equipment aging, system working environment and operating conditions are mainly analyzed and predicted through regular manual inspections and fault records. Therefore, these methods have problems such as reliance on manual inspections, untimely, low accuracy, limitations, data management, analysis difficulties and high costs.

[0003] The fault warning method based on machine learning is a potential fault detection technology that can solve the problems of the above-mentioned traditional warning methods. This technology usually relies on large-scale and high-quality data sets. However, in the actual data set construction process, the collected data is all the data that can be collected in the inverter system. The data types are cumbersome, resulting in excessively large amounts of collected data and uneven quality, which increases the cost of data storage and data processing. At the same time, the nonlinear fitting ability of machine learning is weak, and it is difficult to extract key fault features from existing data, resulting in poor accuracy of fault warning results. Summary of the invention

[0004] The present invention provides a photovoltaic inverter fault warning method, device, equipment and medium, which solves the problem that in the actual warning process, the existing fault warning method lacks sufficient data support due to the small amount of historical fault data. At the same time, the nonlinear fitting ability of machine learning is weak, and it is difficult to extract features associated with the fault from the data set, resulting in poor accuracy of the final warning result.

[0005] In a first aspect, the present invention provides a photovoltaic inverter fault early warning method, comprising the following steps: Conduct correlation analysis on various physical variables of the photovoltaic inverter during operation and different fault types to obtain multiple correlated physical variables; the correlation analysis includes working principle analysis, historical data analysis and correlation experiment analysis; Perform correlation test simulation on multiple associated physical variables and different fault types, and determine multiple candidate physical variables based on the simulation results; Different fault types are set in the photovoltaic inverter simulation system, and fault data corresponding to multiple candidate physical variables are collected. At the same time, normal data of multiple candidate physical variables of the photovoltaic inverter simulation system during normal operation are collected, and a database is constructed based on the fault data and normal data. The convolutional neural network model is trained through the database to obtain a fault warning model; Multiple candidate physical variable data of the photovoltaic inverter during operation are collected in real time and input into the fault warning model to obtain the fault warning result.

[0006] Preferably, the performing correlation analysis on various physical variables during operation of the photovoltaic inverter and different fault types to obtain a plurality of correlated physical variables comprises the following steps: When there are potential faults in the electronic components and circuit structures in the photovoltaic inverter, multiple physical variables are screened based on the working principle of the photovoltaic inverter to obtain a first associated physical variable; Acquire historical operation data of the photovoltaic inverter, locate the position and time of the fault in the historical operation data, extract data of multiple physical variables before the fault occurs, and select multiple physical variables associated with the fault as second associated physical variables; Different fault types are set in the photovoltaic inverter simulation system, multiple physical variables at the output end of the photovoltaic inverter are monitored, and abnormal values ​​of the multiple physical variables are screened to obtain a third associated physical variable; The first associated physical variable, the second associated physical variable and the third associated physical variable are combined to obtain a plurality of associated physical variables.

[0007] Preferably, when collecting the fault data corresponding to a plurality of candidate physical variables, the lowest sampling frequency of each candidate physical variable is selected for sampling.

[0008] Preferably, the building of a database based on fault data and normal data comprises the following steps: According to multiple candidate physical variables and corresponding minimum sampling frequencies, change data of multiple candidate physical variables caused by each fault type are collected as fault data, and the number of samples of each type of fault data is ensured to be the same; Collecting the same number of data of various candidate physical variables of the normal operation of the photovoltaic inverter simulation system as normal data; A database is constructed by using a plurality of fault data and a plurality of normal data.

[0009] Preferably, the convolutional neural network model includes a first convolutional layer, three second convolutional layers, a global average pooling layer, and a fully connected layer, and the training of the convolutional neural network model through the database includes: The fault data of multiple candidate physical variables are processed into time series of equal length by time series expansion and spliced ​​in time order to obtain input variables; The input variables are input into the convolutional neural network model, and the features of the input variables are extracted through the first convolution layer to obtain the first vector; the first vector is transformed into a second vector through three second convolution layers; the second vector is averaged channel by channel through the global average pooling layer to obtain the global feature vector; the global feature vector is output mapped through the fully connected layer to obtain the fault warning result.

[0010] Preferably, it also includes: Build a fault feature library based on fault data; Input multiple candidate physical variable data collected in real time during the operation of the photovoltaic inverter into the fault feature library to obtain the fault detection result; Obtain the average value of the probability value of the same fault in the fault detection result and the fault warning result; The average values ​​of multiple faults are sorted, and the fault type with the largest average value is taken as the warning result.

[0011] Preferably, the plurality of physical variables include DC voltage, DC current, three-phase voltage, three-phase current, power, housing temperature, device temperature and ambient humidity.

[0012] In a second aspect, the present invention provides a photovoltaic inverter fault warning device, comprising: The correlation module is used to perform correlation analysis on various physical variables of the photovoltaic inverter during operation and different fault types to obtain multiple correlated physical variables; wherein the correlation analysis includes working principle analysis, historical data analysis and correlation experiment analysis; A simulation module, used to simulate the correlation test between multiple associated physical variables and different fault types, and determine multiple candidate physical variables based on the simulation results; A construction module is used to set different fault types in the photovoltaic inverter simulation system, collect fault data corresponding to multiple candidate physical variables, and collect normal data of multiple candidate physical variables when the photovoltaic inverter simulation system is in normal operation, and build a database based on the fault data and the normal data; A training module is used to train the convolutional neural network model through a database to obtain a fault warning model; The early warning module is used to collect multiple candidate physical variable data of the photovoltaic inverter during operation in real time, and input it into the fault early warning model to obtain the fault early warning result.

[0013] In a third aspect, the present invention provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned photovoltaic inverter fault warning method when executing the program.

[0014] In a fourth aspect, the present invention provides a computer-readable storage medium, wherein the storage medium stores a computer program, and when the computer program is executed by a processor, the above-mentioned photovoltaic inverter fault warning method is implemented.

[0015] Compared with the prior art, the present invention has the following beneficial effects: The present invention first performs correlation analysis and correlation test on various physical variables of the photovoltaic inverter during operation and different fault types in different aspects. By exploring the correlation between multiple physical variables of the photovoltaic inverter during operation and different fault types, candidate physical variables related to different faults are obtained, and physical variables unrelated to the faults are eliminated, which greatly reduces the types of physical variables and improves data quality. During the training process, the present invention only needs to collect data of candidate physical variables to train the fault warning model, which greatly reduces the scale of data collection and reduces the cost of data storage and data processing. At the same time, the present invention adopts a convolutional neural network model as a fault warning model, which improves the nonlinear fitting ability of data and improves the accuracy of the warning results. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0017] Figure 1 A flowchart of the database and feature library construction scheme of the present invention; Figure 2 It is a schematic diagram of the structure of the fault warning model of the present invention; Figure 3 The figure is a flow chart of the photovoltaic inverter fault early warning method of the present invention. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] In order to address the deficiencies of the prior art, the present invention provides a photovoltaic inverter fault warning method, which uses the communication between the cloud and the device and between the devices to identify and predict possible faults of the photovoltaic inverter in advance, so as to take repair and maintenance measures in time, reduce equipment damage and downtime, and improve the reliability and efficiency of the system. This method first collects electrical data and environmental data in the current photovoltaic inverter system, and selects physical variables closely related to the fault, and then pre-processes the collected physical variables to construct a fault database for the fault detection artificial intelligence model. At the same time, a fault feature library is constructed through the data with faults for fault matching. Based on the constructed fault database and fault feature library, a fault detection model is trained. Various sensors are used to collect the operating data of the photovoltaic inverter. On the one hand, the collected operating data is uploaded, and on the other hand, the fault is identified through the fault warning model and the fault feature library, and finally the detected fault warning signal is uploaded. Reference Figure 1 A photovoltaic inverter fault early warning method of the present invention specifically comprises the following steps: S1: Perform correlation analysis on various physical variables during the operation of the photovoltaic inverter and different fault types to obtain multiple correlated physical variables.

[0020] According to the operating environment of the photovoltaic inverter and the sensors that can be used, the availability of various physical variables is comprehensively analyzed. Physical variables include DC voltage, DC current, three-phase voltage, three-phase current, power, housing temperature, device temperature and ambient humidity.

[0021] The working principle of the photovoltaic inverter is analyzed to obtain the first associated physical variable. According to the electronic components and circuit structure in the photovoltaic inverter, when there are potential faults (parameter change faults, internal open circuits and internal short circuits) in electronic components such as transistors, diodes, capacitors, inductors and resistors, it is determined whether various physical variables will change significantly. Then, when there are short circuits and open circuits in the circuit and when there are extreme temperature and humidity changes in the system operating environment, the data of each physical variable will have significant characteristics.

[0022] The historical data of the photovoltaic inverter is analyzed to obtain the second associated physical variable. According to the historical operation data of the photovoltaic inverter, the location and time of the fault during operation are located, the data of various physical variables before the fault occurs are extracted, and the physical variables that are significantly associated with the fault are selected.

[0023] The simulation software is used to build a photovoltaic inverter simulation system, set fault types such as short circuit, open circuit and overload, monitor the voltage, current and power data at the inverter output, determine the abnormal data, and obtain the third associated physical variable, such as abnormal voltage fluctuation and current drop.

[0024] S2: Perform correlation test simulation (correlation test) on multiple correlated physical variables and different fault types, and determine multiple candidate physical variables based on the simulation results.

[0025] In the photovoltaic inverter simulation system, by manually creating equipment and circuit faults, it is verified whether multiple related physical variables are closely related to the faults, and finally it is determined that various candidate physical variables need to be collected.

[0026] S3: Collect the change data of different candidate physical variables corresponding to the photovoltaic inverter when different faults occur and when it is operating normally, and build a database and fault feature library.

[0027] Use multiple sampling frequencies to collect data of various candidate physical variables, set various fault types in the photovoltaic inverter simulation system, and select the lowest sampling frequency of each candidate physical variable as the established sampling frequency during system operation to reduce data storage space and data processing tasks. In the development stage, each candidate physical variable is collected at multiple sampling frequencies, and then for each candidate physical variable, its availability at different sampling frequencies is verified, and then the lowest sampling frequency that maintains the availability of various candidate physical variables is selected.

[0028] According to the candidate physical variables and data sampling frequency, the change data of various data caused by each potential typical fault is collected as fault data, and the number of samples of each type of fault data is ensured to be the same. Then the same number of data of various candidate physical variables of normal system operation is collected.

[0029] The collected data is cleaned and preprocessed, the time gaps in the collected data are filled in with data backwards, and outliers in the data are removed. Finally, the data is standardized and a database is constructed.

[0030] A database is built for model training, and then the change data of candidate physical variables generated by various typical fault types are integrated into a fault feature library, including various candidate physical variable data and their corresponding labels of whether there is a fault or not.

[0031] S4: Train the convolutional neural network model through the database to obtain a fault warning model.

[0032] The fault data of multiple candidate physical variables are processed into time series of equal length through time series stretching and spliced ​​together in chronological order to obtain input variables (associated variables).

[0033] Input variables are input into the convolutional neural network model. The convolutional neural network model of the present invention is as follows Figure 2 As shown, it includes a first convolution layer, three second convolution layers, a global average pooling layer and a fully connected layer.

[0034] The first and second convolutional layers in the model are both time convolutions. The first convolutional layer increases the dimension of the input variable, and the feature dimension is c , the convolution kernel size is 3, and the sliding step size is set to 2 to reduce the time dimension of the feature. After the output features pass through the activation function ReLU, the maximum pooling layer is used to retain key information to prevent the fault detection model from overfitting. The output result of the first convolution layer is the first vector.

[0035] The sliding step size of the convolution kernel of the second convolution layer is set to 1 to extract high-level feature information. A normalization layer is added after each second convolution layer to keep the order of magnitude of different features consistent, improve the convergence speed of the model, and prevent the fault detection model from overfitting. The output of the three second convolution layers is the second vector.

[0036] After the convolution operation is performed, in order to reduce the dimension of the features, a global average pooling layer is used to obtain the high-level features of the fault variables and obtain the global feature vector.

[0037] The global feature vector is output mapped through the fully connected layer to obtain the fault type. The output prediction part of the model uses a two-layer fully connected neural network to predict whether a fault exists.

[0038] S5: Reference Figure 3 , collect multiple candidate physical variable data of the photovoltaic inverter during operation in real time, and input them into the fault feature library and the fault warning model respectively, to obtain the fault detection results and prediction results, and aggregate the fault detection results and prediction results to obtain the warning signal.

[0039] The fault warning model is configured to the photovoltaic inverter through the Internet of Things. From the photovoltaic inverter, various sensors are used to collect candidate physical variables associated with the fault type, such as voltage, current, power and temperature, according to the data resolution requirements. Specifically, it includes the three-phase AC voltage output by the inverter, the three-phase AC current output by the inverter, the DC current input to the inverter, the DC voltage input to the inverter, the maximum tracking power of the inverter, the internal temperature of the inverter housing, the output three-phase AC voltage frequency, the DC leakage current and the harmonic distortion rate.

[0040] The abnormal data points in the collected data are removed and standardized.

[0041] The standardized data is uploaded to the cloud server based on the Internet of Things technology to record the operating status of the photovoltaic inverter.

[0042] The collected data of various physical variables of the photovoltaic inverter are matched with the fault features in the fault feature library, and the fault detection results of feature matching are output.

[0043] Based on the fault warning model, fault detection is performed on the collected operating data and the prediction results are output.

[0044] In order to make the fault detection result robust and accurate, the prediction results of the two are aggregated to obtain the final fault warning signal. After averaging the probability values ​​of various faults output by the two, the fault type with the largest probability value is selected as the final fault detection result. The warning signal is the fault type detected. The present invention not only greatly reduces the scale of data collection, but also integrates the two warning results of fault detection results and prediction results, thereby improving the accuracy of the warning signal.

[0045] The photovoltaic inverter uploads the generated warning signal to the cloud server based on the Internet of Things technology and finally sends it to the system maintenance unit.

[0046] Based on the same concept, the present invention also provides a photovoltaic inverter fault warning device, including a correlation module, a simulation module, a construction module, a training module and a warning module.

[0047] The correlation module is used to perform correlation analysis on various physical variables of the photovoltaic inverter during operation and different fault types to obtain multiple correlated physical variables; the correlation analysis includes working principle analysis, historical data analysis and correlation experiment analysis.

[0048] The simulation module is used to perform correlation test simulation on multiple associated physical variables and different fault types, and determine multiple candidate physical variables based on the simulation results.

[0049] The construction module is used to set different fault types in the photovoltaic inverter simulation system, collect fault data corresponding to multiple candidate physical variables, and collect normal data of multiple candidate physical variables during normal operation of the photovoltaic inverter simulation system, and build a database based on fault data and normal data.

[0050] The training module is used to train the convolutional neural network model through the database to obtain a fault warning model.

[0051] The early warning module is used to collect multiple candidate physical variable data of the photovoltaic inverter during operation in real time, and input it into the fault early warning model to obtain the fault early warning result.

[0052] The present invention associates the available physical variables with faults by analyzing the working principle of the photovoltaic inverter system, historical operation data and performing association experiments. Physical variables associated with the fault are tested for association to obtain candidate physical variables. Data of candidate physical variables are obtained using different sampling frequencies, a suitable data sampling frequency is selected, photovoltaic inverter fault data and normal data are sampled, data cleaning and preprocessing are performed, and a fault feature library and a database for model training are constructed. A photovoltaic inverter fault warning model based on a convolutional neural network is constructed in the cloud, and the model parameters are optimized by database training, and the optimal model parameters are sent to the inverter system based on the Internet of Things technology. The data of each physical variable of the photovoltaic inverter system is collected and preprocessed by sensors, and the processed data is uploaded to the cloud server. The fault warning model is used to identify whether there is a potential fault in the photovoltaic inverter system. The fault identification result and the fault feature matching result are aggregated, and then the generated warning signal is uploaded to the cloud using the Internet of Things technology.

[0053] The present invention uses the electrical data and environmental data that can be obtained in the current photovoltaic inverter system to select physical variables closely related to the fault, and then pre-processes the collected physical variables based on the Internet of Things technology to build a training model for fault detection artificial intelligence. At the same time, the data with faults is used to build a fault feature library for fault matching. The detected fault warning signal is uploaded based on the Internet of Things technology. The present invention uses the Internet of Things between the cloud and the device and between the devices to identify and predict possible faults of the photovoltaic inverter in advance, so as to take corresponding early warning processing methods such as repair and maintenance measures in time, reduce equipment damage and downtime, and improve the reliability and efficiency of the system.

[0054] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the above-mentioned photovoltaic inverter fault early warning method is implemented when the processor executes the program.

[0055] The present invention also provides a computer-readable storage medium, wherein the storage medium stores a computer program, and when the computer program is executed by a processor, the above-mentioned photovoltaic inverter fault early warning method is implemented.

[0056] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0057] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A photovoltaic inverter fault early warning method, characterized in that: The following steps are involved: Conduct correlation analysis on various physical variables of the photovoltaic inverter during operation and different fault types to obtain multiple correlated physical variables; the correlation analysis includes working principle analysis, historical data analysis and correlation experiment analysis; Perform correlation test simulation on multiple associated physical variables and different fault types, and determine multiple candidate physical variables based on the simulation results; Different fault types are set in the photovoltaic inverter simulation system, and fault data corresponding to multiple candidate physical variables are collected. At the same time, normal data of multiple candidate physical variables of the photovoltaic inverter simulation system during normal operation are collected, and a database is constructed based on the fault data and normal data. The convolutional neural network model is trained through the database to obtain a fault warning model; Multiple candidate physical variable data of the photovoltaic inverter during operation are collected in real time and input into the fault warning model to obtain the fault warning result.

2. A photovoltaic inverter fault early warning method as claimed in claim 1, characterized in that: The process of performing correlation analysis on various physical variables during operation of the photovoltaic inverter and different fault types to obtain a plurality of correlated physical variables includes the following steps: When there are potential faults in the electronic components and circuit structures in the photovoltaic inverter, multiple physical variables are screened based on the working principle of the photovoltaic inverter to obtain a first associated physical variable; Acquire historical operation data of the photovoltaic inverter, locate the position and time of the fault in the historical operation data, extract data of multiple physical variables before the fault occurs, and select multiple physical variables associated with the fault as second associated physical variables; Different fault types are set in the photovoltaic inverter simulation system, multiple physical variables at the output end of the photovoltaic inverter are monitored, and abnormal values ​​of the multiple physical variables are screened to obtain a third associated physical variable; The first associated physical variable, the second associated physical variable and the third associated physical variable are combined to obtain a plurality of associated physical variables.

3. A photovoltaic inverter fault early warning method as claimed in claim 1, characterized in that: When the fault data corresponding to the plurality of candidate physical variables are collected, the lowest sampling frequency of each candidate physical variable is selected for sampling.

4. A photovoltaic inverter fault early warning method as claimed in claim 3, characterized in that: The method of constructing a database based on fault data and normal data includes the following steps: According to multiple candidate physical variables and corresponding minimum sampling frequencies, change data of multiple candidate physical variables caused by each fault type are collected as fault data, and the number of samples of each type of fault data is ensured to be the same; Collecting the same number of data of various candidate physical variables of the normal operation of the photovoltaic inverter simulation system as normal data; A database is constructed by using a plurality of fault data and a plurality of normal data.

5. A photovoltaic inverter fault early warning method as claimed in claim 1, characterized in that: The convolutional neural network model includes a first convolutional layer, three second convolutional layers, a global average pooling layer, and a fully connected layer. The convolutional neural network model is trained by using a database, including: The fault data of multiple candidate physical variables are processed into time series of equal length by time series expansion and spliced ​​in time order to obtain input variables; The input variables are input into the convolutional neural network model, and the features of the input variables are extracted through the first convolution layer to obtain the first vector; the first vector is transformed into a second vector through three second convolution layers; the second vector is averaged channel by channel through the global average pooling layer to obtain the global feature vector; the global feature vector is output mapped through the fully connected layer to obtain the fault warning result.

6. A photovoltaic inverter fault early warning method as claimed in claim 1, characterized in that: Also includes: Build a fault feature library based on fault data; Input multiple candidate physical variable data collected in real time during the operation of the photovoltaic inverter into the fault feature library to obtain the fault detection result; Obtain the average value of the probability value of the same fault in the fault detection result and the fault warning result; The average values ​​of multiple faults are sorted, and the fault type with the largest average value is taken as the warning result.

7. A photovoltaic inverter fault early warning method as claimed in claim 1, characterized in that: The plurality of physical variables include DC voltage, DC current, three-phase voltage, three-phase current, power, housing temperature, device temperature and ambient humidity.

8. A photovoltaic inverter fault warning device, characterized in that: include: The correlation module is used to perform correlation analysis on various physical variables of the photovoltaic inverter during operation and different fault types to obtain multiple correlated physical variables; wherein the correlation analysis includes working principle analysis, historical data analysis and correlation experiment analysis; A simulation module, used to simulate the correlation test between multiple associated physical variables and different fault types, and determine multiple candidate physical variables based on the simulation results; A construction module is used to set different fault types in the photovoltaic inverter simulation system, collect fault data corresponding to multiple candidate physical variables, and collect normal data of multiple candidate physical variables when the photovoltaic inverter simulation system is in normal operation, and build a database based on the fault data and the normal data; A training module is used to train the convolutional neural network model through a database to obtain a fault warning model; The early warning module is used to collect multiple candidate physical variable data of the photovoltaic inverter during operation in real time, and input it into the fault early warning model to obtain the fault early warning result.

9. A computer device, characterized in that: The invention comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the photovoltaic inverter fault early warning method described in any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by the processor, the photovoltaic inverter fault early warning method according to any one of claims 1 to 7 is implemented.

Images