Arc discharge detection method and device of photovoltaic inverter and electronic equipment

Abstract

The invention provides an arc discharge detection method and device for a photovoltaic inverter and electronic equipment, and the method comprises the steps: collecting sample current data corresponding to the photovoltaic inverter, and classifying and marking the sample current data as arc data and arc-free data; taking the sample current data as a data sample, taking the arc data and the arc-free data as sample labels corresponding to the data sample, and training a preset neural network model; converting the trained neural network model into an algorithm program, deploying the algorithm program to a preset micro-control unit, inputting real-time current data collected by the micro-control unit into the neural network model, and determining a data type to which the real-time current data belongs; and when the data type is the electric arc data, determining that the photovoltaic inverter has a direct current arc discharge phenomenon. And low-cost and high-precision intelligent arc discharge detection can be realized.

Description

Technical Field

[0001] The present disclosure relates to the technical field of inverters, and in particular to an arc detection method, device and electronic equipment for a photovoltaic inverter. Background Art

[0002] With the rapid development and popularization of photovoltaic power generation technology, the application of photovoltaic systems has shown explosive growth worldwide. However, since photovoltaic power stations use direct current as the main form of electricity, DC arcs have become one of the biggest safety hazards in photovoltaic systems. Statistics show that DC arcing is the main cause of fire accidents in photovoltaic power stations. How to achieve accurate and efficient arc detection and respond quickly when an arc is detected has become a technical problem that photovoltaic equipment manufacturers and system integrators need to solve urgently.

[0003] Existing arc detection methods mainly rely on detection technologies based on digital signal processors (DSPs) or dedicated function modules. These methods judge the arc state by analyzing the frequency, energy, and change of current or voltage signals, and use fixed thresholds to debug parameters in actual scenarios. However, traditional detection methods are limited by the complexity of the algorithm model and cannot fully capture complex arc characteristics, especially in environmental noise interference and diversified application scenarios, and the detection results are less accurate. Existing technologies have delays in signal processing and judgment, and may not be able to respond to arc status in a timely manner, increasing safety hazards. In addition, solutions based on DSPs or dedicated function modules have high requirements for hardware resources, resulting in increased system costs and difficulty in promoting applications in large-scale photovoltaic systems. Summary of the invention

[0004] The embodiments of the present disclosure at least provide a method, device and electronic equipment for arc detection of a photovoltaic inverter, which can realize low-cost and high-precision intelligent arc detection.

[0005] The present disclosure provides an arc detection method for a photovoltaic inverter, comprising:

[0006] Collecting sample current data corresponding to the photovoltaic inverter, and classifying and marking the sample current data into arc data and no-arc data;

[0007] The sample current data is used as a data sample, and the arc data and the no-arc data are used as sample labels corresponding to the data sample to train a preset neural network model;

[0008] Convert the trained neural network model into an algorithm program and deploy it to a preset microcontroller unit, input the real-time current data collected by the microcontroller unit into the neural network model, and determine the data type to which the real-time current data belongs;

[0009] When the data type is the arc data, it is determined that a DC arcing phenomenon occurs in the photovoltaic inverter.

[0010] In an optional implementation manner, collecting sample current data corresponding to the photovoltaic inverter specifically includes:

[0011] In the preset simulation environment and the actual use environment, artificial arc simulation is performed through the arc base;

[0012] The sample current data under different working conditions are collected in real time by the micro control unit;

[0013] The sample current data is transmitted to a preset data acquisition host computer via a high-speed USB interface;

[0014] The data acquisition host computer classifies the sample current data into the arc data and the non-arc data, and marks the sample current data with data labels of corresponding categories.

[0015] In an optional implementation manner, after classifying and marking the sample current data as arc data and no-arc data, the method further includes:

[0016] Inputting the sample current data into a preset signal analysis host computer;

[0017] Analyze the frequency domain characteristics and time domain characteristics corresponding to the sample current data through the signal analysis host computer;

[0018] Valid sample data are screened out from the sample current data according to differences between the arc data and the non-arc data corresponding to the frequency domain features and the time domain features.

[0019] In an optional implementation manner, when the data type is the arc data, after determining that a DC arcing phenomenon occurs in the photovoltaic inverter, the method further includes:

[0020] A serial communication interface protocol is used to establish a communication channel between the main controller corresponding to the photovoltaic inverter;

[0021] The DC arcing fault information is reported to the main controller via the communication channel.

[0022] In an optional implementation, after converting the trained neural network model into an algorithm program and deploying it to a preset micro control unit, the method further includes:

[0023] In a preset simulation environment and an actual use environment, artificial arc drawing is simulated by an arc drawing base, and verification current data corresponding to the photovoltaic inverter under different working conditions is collected in real time by the micro control unit;

[0024] According to the data type corresponding to the verification current data, the Yang Zheng current data tag includes the data labels of the arc data and the arc-free data, generating a verification data set;

[0025] Inputting the verification data set into the trained neural network model to verify the classification accuracy and response speed of the neural network model;

[0026] The model parameters of the neural network model are adjusted according to the verification results so that the neural network model meets the preset operation requirements.

[0027] The present disclosure also provides an arc detection device for a photovoltaic inverter, the device comprising:

[0028] A data acquisition module, used for acquiring sample current data corresponding to the photovoltaic inverter, and classifying and marking the sample current data into arc data and no-arc data;

[0029] A model training module, used to train a preset neural network model using the sample current data as data samples, the arc data and the no-arc data as sample labels corresponding to the data samples;

[0030] A model deployment module, used to convert the trained neural network model into an algorithm program and deploy it to a preset micro control unit, input the real-time current data collected by the micro control unit into the neural network model, and determine the data type to which the real-time current data belongs;

[0031] The arcing detection module is used to determine that a DC arcing phenomenon occurs in the photovoltaic inverter when the data type is the arcing data.

[0032] In an optional implementation manner, the data acquisition module is specifically used for:

[0033] In the preset simulation environment and the actual use environment, artificial arc simulation is performed through the arc base;

[0034] The sample current data under different working conditions are collected in real time by the micro control unit;

[0035] The sample current data is transmitted to a preset data acquisition host computer via a high-speed USB interface;

[0036] The data acquisition host computer classifies the sample current data into the arc data and the non-arc data, and marks the sample current data with data labels of corresponding categories.

[0037] In an optional embodiment, the device is also used for:

[0038] Inputting the sample current data into a preset signal analysis host computer;

[0039] Analyze the frequency domain characteristics and time domain characteristics corresponding to the sample current data through the signal analysis host computer;

[0040] Valid sample data are screened out from the sample current data according to differences between the arc data and the non-arc data corresponding to the frequency domain features and the time domain features.

[0041] The embodiment of the present disclosure also provides an electronic device, including: a processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate through the bus, and when the machine-readable instructions are executed by the processor, the above-mentioned arc detection method for the photovoltaic inverter, or the steps in any possible implementation of the above-mentioned arc detection method for the photovoltaic inverter are executed.

[0042] The embodiments of the present disclosure further provide a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the computer program executes the arc detection method for the photovoltaic inverter or the steps in any possible implementation of the arc detection method for the photovoltaic inverter.

[0043] The embodiments of the present disclosure also provide a computer program product, including a computer program / instruction, which, when executed by a processor, implements the above-mentioned arc detection method for the photovoltaic inverter, or the steps in any possible implementation of the above-mentioned arc detection method for the photovoltaic inverter.

[0044] The embodiment of the present disclosure provides a photovoltaic inverter arc detection method, device and electronic device, which collects sample current data corresponding to the photovoltaic inverter, and classifies and marks the sample current data as arc data and no-arc data; uses the sample current data as data samples, and the arc data and no-arc data as sample labels corresponding to the data samples to train a preset neural network model; converts the trained neural network model into an algorithm program and deploys it to a preset microcontroller unit, inputs the real-time current data collected by the microcontroller unit into the neural network model, and determines the data type to which the real-time current data belongs; when the data type is the arc data, it is determined that the photovoltaic inverter has a DC arc phenomenon. Low-cost and high-precision intelligent arc detection can be achieved.

[0045] In order to make the above-mentioned objectives, features and advantages of the present disclosure more obvious and easy to understand, preferred embodiments are specifically cited below and described in detail with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the following is a brief introduction to the drawings required for use in the embodiments. The drawings herein are incorporated into the specification and constitute a part of the specification. These drawings illustrate embodiments consistent with the present disclosure and are used together with the specification to illustrate the technical solutions of the present disclosure. It should be understood that the following drawings only illustrate certain embodiments of the present disclosure and should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can also be obtained based on these drawings without creative work.

[0047] Figure 1 A flow chart of an arc detection method for a photovoltaic inverter provided by an embodiment of the present disclosure is shown;

[0048] Figure 2 A flow chart of another method for detecting arcing of a photovoltaic inverter provided by an embodiment of the present disclosure is shown;

[0049] Figure 3 A schematic diagram of an arc detection device for a photovoltaic inverter provided by an embodiment of the present disclosure is shown;

[0050] Figure 4 A schematic diagram of an electronic device provided by an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical scheme and advantages of the embodiments of the present disclosure clearer, the technical scheme in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only part of the embodiments of the present disclosure, rather than all of the embodiments. The components of the embodiments of the present disclosure generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present disclosure provided in the drawings is not intended to limit the scope of the present disclosure for protection, but merely represents the selected embodiments of the present disclosure. Based on the embodiments of the present disclosure, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present disclosure.

[0052] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0053] The term "and / or" herein only describes an association relationship, indicating that three relationships may exist. For example, A and / or B may represent the following three situations: A exists alone, A and B exist at the same time, and B exists alone. In addition, the term "at least one" herein represents any combination of at least two of any one or more of a plurality of. For example, including at least one of A, B, and C may represent including any one or more elements selected from the set consisting of A, B, and C.

[0054] Research has found that existing arc detection methods mainly rely on detection technologies based on digital signal processors (DSPs) or dedicated function modules. These methods judge the arc state by analyzing the frequency, energy, and change amount of current or voltage signals, and use fixed thresholds to debug parameters in actual scenarios. However, traditional detection methods are limited by the complexity of the algorithm model and cannot fully capture complex arc characteristics, especially in environmental noise interference and diversified application scenarios, and the detection results are less accurate. The existing technology has delays in the signal processing and judgment process, and may not be able to respond to the arc state in time, increasing safety hazards. In addition, solutions based on DSP or dedicated function modules have high requirements for hardware resources, resulting in increased system costs and difficulty in promoting and applying them in large-scale photovoltaic systems.

[0055] Based on the above research, the present disclosure provides a method, device and electronic device for detecting arcing of a photovoltaic inverter, which collects sample current data corresponding to the photovoltaic inverter, and classifies and marks the sample current data as arcing data and no-arc data; uses the sample current data as data samples, and the arcing data and no-arc data as sample labels corresponding to the data samples, and trains a preset neural network model; converts the trained neural network model into an algorithm program and deploys it to a preset microcontroller unit, inputs the real-time current data collected by the microcontroller unit into the neural network model, and determines the data type to which the real-time current data belongs; when the data type is the arcing data, it is determined that the photovoltaic inverter has a DC arcing phenomenon. Low-cost and high-precision intelligent arcing detection can be achieved.

[0056] To facilitate understanding of this embodiment, a photovoltaic inverter arc detection method disclosed in the embodiment of the present disclosure is first introduced in detail. The execution subject of the photovoltaic inverter arc detection method provided in the embodiment of the present disclosure is generally a computer device with certain computing capabilities, and the computer device includes, for example: a terminal device or a server or other processing device, and the terminal device can be a user equipment (User Equipment, UE), a mobile device, a user terminal, a terminal, a cellular phone, a cordless phone, a personal digital assistant (Personal Digital Assistant, PDA), a handheld device, a computing device, a vehicle-mounted device, a wearable device, etc. In some possible implementations, the photovoltaic inverter arc detection method can be implemented by a processor calling a computer-readable instruction stored in a memory.

[0057] See also Figure 1 FIG. 1 is a flow chart of a method for detecting arcing of a photovoltaic inverter provided by an embodiment of the present disclosure, the method comprising steps S101 to S104, wherein:

[0058] S101 , collecting sample current data corresponding to a photovoltaic inverter, and classifying and marking the sample current data into arc data and no-arc data.

[0059] In the specific implementation, artificial arc drawing is simulated through an arc drawing base in a preset simulation environment and an actual use environment; sample current data under different working conditions are collected in real time through a microcontroller unit; the sample current data is transmitted to a preset data acquisition host computer through a high-speed USB interface; the sample current data is classified into arc data and no-arc data through the data acquisition host computer, and data labels of corresponding categories are marked for the sample current data.

[0060] Here, an arc drawing machine is used to simulate the generation process of a DC arc in a laboratory simulation scenario or in an actual use environment. The microcontroller unit is responsible for collecting current signals. The current sensor is used to capture the current signal of the inverter to ensure that the signal is accurate. A high-speed data transmission interface such as a USB interface is used to transmit the collected data to a host computer.

[0061] Among them, according to the common operating conditions of photovoltaic inverters, a variety of operating conditions are designed, including normal operating conditions (no arcing); abnormal conditions (obvious arcing phenomenon); operation under different power outputs, load conditions and environmental noise interference.

[0062] Specifically, an arc machine is used to simulate arcing in the DC circuit to trigger arc signals of different intensities; at the same time, the current signal under normal operation is collected as arc-free data. In the actual operating environment of the photovoltaic system, the current signal of the inverter during normal operation is collected to ensure that the data has real scene characteristics; possible faults (such as poor contact at the connection point or aging of the line) are simulated to induce actual arc phenomena and collect corresponding current signals.

[0063] Here, the sampling frequency is set according to the operating frequency of the inverter to ensure that the high-frequency characteristics of the arc signal are captured; the current signals under various scenarios and working conditions are collected to ensure the breadth and representativeness of the data. The collected raw current data is transmitted to the host computer through a high-speed USB interface.

[0064] Furthermore, the collected current data is denoised to filter out environmental noise and interference signals from the collection equipment; the data is sliced ​​according to the time window to form sample data blocks of fixed length, and the current data in the corresponding time window is marked as "arc" according to the arc drawing experiment record or the time point when the arc is actually induced; the current data collected during the period of stable operation of the system is marked as "arc-free" to avoid selecting data segments that may have environmental interference or sudden current fluctuations, thereby ensuring the accuracy of the marking.

[0065] Here, the labeled current data is stored in a standardized format (such as CSV or JSON), and the labeled data is classified and stored according to whether there is an arc or no arc, which is convenient for subsequent signal analysis and model training.

[0066] As a possible implementation manner, after classifying and marking the sample current data as arc data and no-arc data, the sample current data is input into a preset signal analysis host computer; the frequency domain characteristics and time domain characteristics corresponding to the sample current data are analyzed by the signal analysis host computer; and based on the differences in the frequency domain characteristics and the time domain characteristics corresponding to the arc data and the no-arc data, valid sample data is screened out from the sample current data.

[0067] In the specific implementation, the signal analysis host computer is a tool for analyzing current data, with functions such as time domain waveform drawing, frequency domain feature extraction, data comparison and screening. The tool presets the frequency domain and time domain analysis algorithms required for arc detection, such as fast Fourier transform (FFT) and waveform change detection. The classified and marked current data is uploaded to the signal analysis host computer in a standard format (such as CSV or binary file); the system automatically identifies the data type and loads it according to the arc and no arc classification to ensure clear data grouping.

[0068] Here, by analyzing the original waveform data of the current signal, key time domain features are extracted. The time domain features include but are not limited to the current fluctuation amplitude, peak value and average value, waveform fluctuation rate, etc. The time domain features of the arc data and the non-arc data are compared. The waveform fluctuation of the arc data is obviously more violent and has irregular sudden fluctuations; the waveform of the non-arc data is relatively stable, with low fluctuation frequency and small amplitude change.

[0069] Here, the fast Fourier transform (FFT) algorithm is used to convert the time domain signal into a frequency domain signal to extract the frequency characteristics of the current signal. Arc signals usually cause an abnormal increase in energy in a specific frequency band; the frequency domain distribution energy of the arc-free data is concentrated and relatively smooth. The frequency component with the largest energy in the signal is extracted, and the spectrum distribution range of the signal is calculated to statistically analyze the energy value changes in a specific frequency band (such as the high-frequency area). The characteristic differences between the data with and without arc are compared through the frequency domain graph. The energy peak of the data with arc in the high frequency band (such as above 1kHz) is significantly higher than that of the data without arc; the energy of the data without arc in the low frequency band (such as below 100Hz) is more concentrated.

[0070] Furthermore, based on the analysis results, the main difference rules between data with arc and data without arc in the frequency domain and time domain are defined. For example, samples with a fluctuation amplitude in the time domain greater than a certain threshold are marked as valid data with arc; samples with high-frequency energy in the frequency domain exceeding a certain range are marked as valid data with arc; and data without arc that meets the above rules are marked as valid data without arc.

[0071] Here, the system automatically matches data with difference rules to preliminarily screen qualified samples; compares the preliminary screening results with the manually marked classifications to eliminate misclassified samples; and manually reviews invalid or ambiguous samples to ensure the accuracy of the screening results. The screened valid sample data includes: high-quality data with obvious arc characteristics (valid arc samples) and reference data under stable operating conditions (valid arc-free samples).

[0072] In this way, high-quality, representative and valid samples can be effectively screened out from the collected sample current data, providing a solid data foundation for subsequent AI model training and arc detection algorithm optimization.

[0073] S102, using the sample current data as data samples, and the arc data and the no-arc data as sample labels corresponding to the data samples, to train a preset neural network model.

[0074] In the specific implementation, after completing the classification labeling of the sample current data and the screening of valid data, the sample current data is used as the input data sample, and the arc data and no-arc data are used as sample labels to train the preset neural network model and generate an intelligent algorithm with arc detection capability.

[0075] Here, the sample current data is filtered to form a high-quality input data set. Each sample data segment is usually composed of a fixed-length current value sequence (such as the current sampling value within a time window). The sample label of arc data is set to "1", indicating that an arc occurs; the sample label of no arc data is set to "0", indicating that no arc occurs. The label provides a supervisory signal for model training, guiding the model to learn the association between input data and arc status.

[0076] Among them, the sample data is divided into a training set (about 70%-80%), a validation set (about 10%-20%) and a test set (about 10%) for model training, parameter tuning and effect evaluation.

[0077] Furthermore, according to the timing characteristics of the current signal, a suitable neural network structure is selected, such as a convolutional neural network (CNN) and a recurrent neural network (RNN) or a long short-term memory network (LSTM). The convolutional neural network extracts local features in the sample data, such as waveform features or frequency domain features; the recurrent neural network and the long short-term memory network capture the time series characteristics in the sample data, such as the waveform change trend.

[0078] Here, in the neural network model, the input layer accepts sample current data as input, and each sample corresponds to a current sequence of fixed length; the feature extraction layer uses a convolutional layer to extract local features of the current data (such as frequency domain and waveform features); the LSTM layer is used to capture the timing characteristics of the current data (such as fluctuation trends); the fully connected layer integrates the extracted features and outputs the classification results; the output layer uses a binary classification output (such as arc occurrence and no arc occurrence), corresponding to the sample label.

[0079] Among them, the cross entropy loss function is used to measure the error between the model prediction results and the actual labels.

[0080] Here, the sample current data in the training set is input into the neural network, and the corresponding sample labels are used as supervisory signals to guide the model to adjust parameters. The neural network extracts the features of the input data and predicts the labels through forward propagation. The model parameters are updated by calculating the error between the predicted value and the true label. Multiple rounds of iterations are repeated until the model reaches the desired classification accuracy.

[0081] Preferably, after each round of training, use the validation set to evaluate the model performance and adjust the hyperparameters (such as learning rate, number of network layers, number of nodes, etc.) to optimize the model effect. Use regularization techniques (such as Dropout or L2 regularization) to reduce the risk of overfitting, and use data enhancement (such as noise addition or time series slicing) to improve the generalization ability of the model.

[0082] As a possible implementation method, after the neural network model training is completed, artificial arcing is simulated through an arcing base in a preset simulation environment and an actual use environment, and the verification current data corresponding to the photovoltaic inverter under different working conditions is collected in real time through a microcontroller unit; according to the data type corresponding to the verification current data, the current data is labeled as Yang Zheng, including data labels with arc data and without arc data, to generate a verification data set; the verification data set is input into the trained neural network model, and the classification accuracy and response speed of the neural network model are verified; the model parameters of the neural network model are adjusted according to the verification results, so that the neural network model meets the preset operating requirements.

[0083] In the specific implementation, the test set is used to evaluate the performance of the model on unknown data, and indicators such as accuracy, precision, recall, and F1 score are calculated. The classification results of the model for samples with and without arcs are observed through the confusion matrix, and the misclassified samples are analyzed to optimize the training data or model structure.

[0084] S103, converting the trained neural network model into an algorithm program and deploying it to a preset micro control unit, inputting the real-time current data collected by the micro control unit into the neural network model, and determining the data type to which the real-time current data belongs.

[0085] S104: When the data type is the arc data, determine that a DC arcing phenomenon occurs in the photovoltaic inverter.

[0086] In the specific implementation, the trained neural network model is saved as a model file (such as ONNX or TensorFlowLite format) for subsequent deployment. The file contains the model's weight parameters and network structure. The model is quantized (such as converting floating point numbers to integers) or pruned (removing redundant nodes) to reduce the model size and inference time to adapt to the resource limitations of embedded devices.

[0087] Here, the trained model file is integrated with the algorithm code to generate a program that can run on the embedded MCU; the running effect of the deployed version is verified with test data, including detection accuracy and response speed.

[0088] Among them, use embedded tools (such as CMSIS-NN, TensorFlow Lite Micro) to convert the model file into C language or other code forms adapted to the MCU to generate an algorithm program that can run in the MCU. Integrate the converted algorithm program with the embedded software of the MCU, compile the program, download it to the internal storage of the MCU (such as Flash), configure the startup code of the MCU, and ensure that the algorithm is correctly loaded and run after the system is initialized.

[0089] In the specific application process, the real-time collected current data fragments are input into the neural network model deployed in the MCU. The MCU performs forward propagation operations and calculates the data classification results based on the model weights and network structure. According to the classification results and confidence thresholds (such as 0.9) output by the model, the type of the current data is determined. If there is arc data, further protection measures (such as circuit breaker and alarm) are triggered. If there is no arc data, it runs normally without triggering additional operations.

[0090] The embodiment of the present disclosure provides a method for detecting arcing of a photovoltaic inverter, which collects sample current data corresponding to the photovoltaic inverter, and classifies and marks the sample current data as arcing data and no-arc data; uses the sample current data as data samples, and the arcing data and no-arc data as sample labels corresponding to the data samples, and trains a preset neural network model; converts the trained neural network model into an algorithm program and deploys it to a preset microcontroller unit, inputs the real-time current data collected by the microcontroller unit into the neural network model, and determines the data type to which the real-time current data belongs; when the data type is the arcing data, it is determined that the photovoltaic inverter has a DC arcing phenomenon. Low-cost and high-precision intelligent arcing detection can be achieved.

[0091] See also Figure 2 As shown, it is a flow chart of another method for detecting arcing of a photovoltaic inverter provided by an embodiment of the present disclosure, the method comprising steps S201 to S205, wherein:

[0092] S201 , collecting sample current data corresponding to a photovoltaic inverter, and classifying and marking the sample current data into arc data and no-arc data.

[0093] S202, using the sample current data as data samples, and the arc data and the no-arc data as sample labels corresponding to the data samples, to train a preset neural network model.

[0094] S203, converting the trained neural network model into an algorithm program and deploying it to a preset micro control unit, inputting the real-time current data collected by the micro control unit into the neural network model, and determining the data type to which the real-time current data belongs.

[0095] S204: When the data type is the arc data, determine that a DC arcing phenomenon occurs in the photovoltaic inverter.

[0096] It should be noted that steps S201 to S204 are substantially the same as steps S101 to S104, have the same implementation methods and can achieve the same technical effects, and will not be described in detail.

[0097] S205, establishing a communication channel between a main controller corresponding to the photovoltaic inverter and the main controller by using a serial communication interface protocol; and reporting DC arc fault information to the main controller through the communication channel.

[0098] In a specific implementation, when the real-time data is classified as "arc present", the MCU can report information to the inverter main control system through the communication interface. The reported content may include the detected arc fault status, the specific channel where the arc occurs (such as a certain circuit of the photovoltaic array), and the detection time and confidence information.

[0099] Here, the SCI interface custom protocol is used to communicate with the inverter master control to report the detected arc fault information, including the fault channel and specific location.

[0100] Optionally, you can use the 485 communication protocol to receive instructions sent by the header, support issuing upgrade commands, and implement remote firmware upgrades of the module. Use the Ymodem protocol to implement firmware upgrades through 485 communication, transmit the new firmware to the MCU, complete remote upgrades of the module, and ensure that the device functions are always in the best state.

[0101] The embodiment of the present disclosure provides a method for detecting arcing of a photovoltaic inverter, which collects sample current data corresponding to the photovoltaic inverter, and classifies and marks the sample current data as arcing data and no-arc data; uses the sample current data as data samples, and the arcing data and no-arc data as sample labels corresponding to the data samples to train a preset neural network model; converts the trained neural network model into an algorithm program and deploys it to a preset microcontroller unit, inputs the real-time current data collected by the microcontroller unit into the neural network model, and determines the data type to which the real-time current data belongs; when the data type is the arcing data, it is determined that the photovoltaic inverter has a DC arcing phenomenon. A serial communication interface protocol is used to establish a communication channel between the main controller corresponding to the photovoltaic inverter; and the DC arcing fault information is reported to the main controller through the communication channel. Low-cost and high-precision intelligent arcing detection can be achieved.

[0102] Those skilled in the art will appreciate that, in the above method of specific implementation, the order in which the steps are written does not imply a strict execution order and does not constitute any limitation on the implementation process. The specific execution order of the steps should be determined by their functions and possible internal logic.

[0103] Based on the same inventive concept, an arc detection device for a photovoltaic inverter corresponding to the arc detection method for a photovoltaic inverter is also provided in the embodiment of the present disclosure. Since the principle of solving the problem by the device in the embodiment of the present disclosure is similar to the arc detection method for the photovoltaic inverter mentioned in the embodiment of the present disclosure, the implementation of the device can refer to the implementation of the method, and the repeated parts will not be repeated.

[0104] See also Figure 3 , Figure 3 A schematic diagram of an arc detection device for a photovoltaic inverter provided by an embodiment of the present disclosure. Figure 3 As shown in the figure, the arc detection device 300 of the photovoltaic inverter provided by the embodiment of the present disclosure includes:

[0105] The data acquisition module 310 is used to acquire sample current data corresponding to the photovoltaic inverter, and classify and mark the sample current data into arc data and no-arc data.

[0106] The model training module 320 is used to train a preset neural network model using the sample current data as data samples, and the arc data and the no-arc data as sample labels corresponding to the data samples.

[0107] The model deployment module 330 is used to convert the trained neural network model into an algorithm program and deploy it to a preset micro control unit, input the real-time current data collected by the micro control unit into the neural network model, and determine the data type to which the real-time current data belongs.

[0108] The arc detection module 340 is used to determine whether a DC arc phenomenon occurs in the photovoltaic inverter when the data type is the arc data.

[0109] For descriptions of the processing flow of each module in the device and the interaction flow between each module, reference may be made to the relevant descriptions in the above method embodiment, which will not be described in detail here.

[0110] The embodiment of the present disclosure provides an arc detection device for a photovoltaic inverter, which collects sample current data corresponding to the photovoltaic inverter, and classifies and marks the sample current data into arc data and no-arc data; uses the sample current data as data samples, and the arc data and no-arc data as sample labels corresponding to the data samples to train a preset neural network model; converts the trained neural network model into an algorithm program and deploys it to a preset microcontroller unit, inputs the real-time current data collected by the microcontroller unit into the neural network model, and determines the data type to which the real-time current data belongs; when the data type is the arc data, it is determined that the photovoltaic inverter has a DC arc phenomenon. A serial communication interface protocol is used to establish a communication channel between the main controller corresponding to the photovoltaic inverter; and the DC arc fault information is reported to the main controller through the communication channel. Low-cost, high-precision intelligent arc detection can be achieved.

[0111] Corresponds to Figure 1 The arc detection method of the photovoltaic inverter in the present disclosure also provides an electronic device 400, such as Figure 4 FIG. 4 is a schematic diagram of the structure of an electronic device 400 provided in an embodiment of the present disclosure, including:

[0112] Processor 41, memory 42, and bus 43; memory 42 is used to store execution instructions, including memory 421 and external memory 422; memory 421 here is also called internal memory, which is used to temporarily store operation data in processor 41 and data exchanged with external memory 422 such as hard disk. Processor 41 exchanges data with external memory 422 through memory 421. When the electronic device 400 is running, the processor 41 communicates with the memory 42 through bus 43, so that the processor 41 executes Figure 1 and Figure 2 The steps of the arc detection method of the photovoltaic inverter.

[0113] The present disclosure also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the photovoltaic inverter arc detection method described in the above method embodiment are executed. The storage medium can be a volatile or non-volatile computer-readable storage medium.

[0114] The present disclosure also provides a computer program product, which includes computer instructions. When the computer instructions are executed by a processor, the steps of the arc detection method for the photovoltaic inverter described in the above method embodiment can be executed. For details, please refer to the above method embodiment, which will not be repeated here.

[0115] The computer program product may be implemented in hardware, software or a combination thereof. In one optional embodiment, the computer program product is implemented as a computer storage medium. In another optional embodiment, the computer program product is implemented as a software product, such as a software development kit (SDK).

[0116] Those skilled in the art can clearly understand that, for the convenience and simplicity of description, the specific working process of the device described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here. In the several embodiments provided in the present disclosure, it should be understood that the disclosed device and method can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of the device or unit can be electrical, mechanical or other forms.

[0117] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0118] In addition, each functional unit in each embodiment of the present disclosure may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0119] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium that is executable by a processor. Based on this understanding, the technical solution of the present disclosure, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present disclosure. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0120] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present disclosure, which are used to illustrate the technical solutions of the present disclosure, rather than to limit them. The protection scope of the present disclosure is not limited thereto. Although the present disclosure is described in detail with reference to the above-described embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-described embodiments within the technical scope disclosed in the present disclosure, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present disclosure, and should be included in the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure shall be subject to the protection scope of the claims.

Claims

1. A method for detecting arcing of a photovoltaic inverter, characterized in that: include: Collecting sample current data corresponding to the photovoltaic inverter, and classifying and marking the sample current data into arc data and no-arc data; The sample current data is used as a data sample, and the arc data and the no-arc data are used as sample labels corresponding to the data sample to train a preset neural network model; Convert the trained neural network model into an algorithm program and deploy it to a preset microcontroller unit, input the real-time current data collected by the microcontroller unit into the neural network model, and determine the data type to which the real-time current data belongs; When the data type is the arc data, it is determined that a DC arcing phenomenon occurs in the photovoltaic inverter.

2. The method according to claim 1, characterized in that Collect sample current data corresponding to the photovoltaic inverter, including: In the preset simulation environment and the actual use environment, artificial arc simulation is performed through the arc base; The sample current data under different working conditions are collected in real time by the micro control unit; The sample current data is transmitted to a preset data acquisition host computer via a high-speed USB interface; The data acquisition host computer classifies the sample current data into the arc data and the non-arc data, and marks the sample current data with data labels of corresponding categories.

3. The method according to claim 1, characterized in that After classifying and marking the sample current data as arc data and no-arc data, the method further includes: Inputting the sample current data into a preset signal analysis host computer; Analyze the frequency domain characteristics and time domain characteristics corresponding to the sample current data through the signal analysis host computer; Valid sample data are screened out from the sample current data according to differences between the arc data and the non-arc data corresponding to the frequency domain features and the time domain features.

4. The method according to claim 1, characterized in that When the data type is the arc data, after determining that a DC arcing phenomenon occurs in the photovoltaic inverter, the method further includes: A serial communication interface protocol is used to establish a communication channel between the main controller corresponding to the photovoltaic inverter; The DC arcing fault information is reported to the main controller via the communication channel.

5. The method according to claim 1, characterized in that After converting the trained neural network model into an algorithm program and deploying it to a preset micro control unit, the method further includes: In a preset simulation environment and an actual use environment, artificial arc drawing is simulated by an arc drawing base, and verification current data corresponding to the photovoltaic inverter under different working conditions is collected in real time by the micro control unit; According to the data type corresponding to the verification current data, the Yang Zheng current data tag includes the data labels of the arc data and the arc-free data, generating a verification data set; Inputting the verification data set into the trained neural network model to verify the classification accuracy and response speed of the neural network model; The model parameters of the neural network model are adjusted according to the verification results so that the neural network model meets the preset operation requirements.

6. An arc detection device for a photovoltaic inverter, characterized in that: include: A data acquisition module, used for acquiring sample current data corresponding to the photovoltaic inverter, and classifying and marking the sample current data into arc data and no-arc data; A model training module, used to train a preset neural network model using the sample current data as data samples, the arc data and the no-arc data as sample labels corresponding to the data samples; A model deployment module, used to convert the trained neural network model into an algorithm program and deploy it to a preset micro control unit, input the real-time current data collected by the micro control unit into the neural network model, and determine the data type to which the real-time current data belongs; The arcing detection module is used to determine that a DC arcing phenomenon occurs in the photovoltaic inverter when the data type is the arcing data.

7. The device according to claim 6, characterized in that The data acquisition module is specifically used for: In the preset simulation environment and the actual use environment, artificial arc simulation is performed through the arc base; The sample current data under different working conditions are collected in real time by the micro control unit; The sample current data is transmitted to a preset data acquisition host computer via a high-speed USB interface; The data acquisition host computer classifies the sample current data into the arc data and the non-arc data, and marks the sample current data with data labels of corresponding categories.

8. The device according to claim 6, characterized in that The device is also used for: Inputting the sample current data into a preset signal analysis host computer; Analyze the frequency domain characteristics and time domain characteristics corresponding to the sample current data through the signal analysis host computer; Valid sample data are screened out from the sample current data according to differences between the arc data and the non-arc data corresponding to the frequency domain features and the time domain features.

9. An electronic device, characterized in that: include: A processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate via the bus, and when the machine-readable instructions are executed by the processor, the steps of the arc detection method for a photovoltaic inverter as described in any one of claims 1 to 5 are performed.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the arc detection method for a photovoltaic inverter as described in any one of claims 1 to 5 are executed.

Images