Abnormality detection method for distributed power supply access unit

Abstract

The invention provides an anomaly detection method for a distributed power supply access unit, and relates to the technical field of photovoltaic equipment detection, and the method comprises the steps: building a simulation model of a photovoltaic inverter in advance; wherein the simulation model comprises at least two types of photovoltaic inverters, and each photovoltaic inverter corresponds to one communication protocol; reading the simulation model by using a distributed power supply access unit; for each photovoltaic inverter, testing the inverter data of the distributed power supply access unit and the on-off control of the inverter based on the consistency of the photovoltaic inverter data read from the simulation model by using the distributed power supply access unit and the corresponding photovoltaic inverter data directly read by the upper computer; and determining the abnormal state of the distributed power supply access unit according to the test result of the distributed power supply access unit for the inverter data of the various types of photovoltaic inverters and the inverter startup and shutdown control. According to the scheme, the anomaly detection cost and the system complexity of the distributed power supply access unit can be reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic equipment detection, and in particular to an abnormality detection method for a distributed power supply access unit. Background Art

[0002] As the global demand for renewable energy continues to grow, photovoltaic power generation is rapidly expanding in scope and scale as a clean and sustainable energy solution. As a key component in photovoltaic power generation systems, photovoltaic inverters are responsible for converting the direct current generated by solar panels into alternating current for use in the power grid or directly to end users. However, due to the existence of many photovoltaic inverter manufacturers on the market, each manufacturer may use different communication protocols to enable the inverter to interact with the power grid or other smart devices, which leads to the diversity and complexity of the protocol.

[0003] This diversity of protocols is particularly prominent in distributed power access units. Distributed power access units need to be able to communicate effectively with photovoltaic inverters of multiple different protocols to monitor, control and optimize power generation. However, since the communication protocols used or compatible with the inverter are complex and diverse, it is easy to cause abnormalities in the distributed power access unit. That is, in practice, there may be distributed power access units that cannot fully meet and adapt to the various communication protocols of the connected inverter system, which may make it difficult for the inverter to communicate with the power grid. Therefore, it is necessary to detect the communication conversion function of the distributed power access unit to ensure that the distributed access unit can perform communication conversion normally and accurately, thereby ensuring the stable operation and high efficiency of the photovoltaic system.

[0004] Traditionally, abnormal detection of distributed power access units often relies on customizing the corresponding hardware for each communication protocol. However, since the inverter is compatible and applicable to a variety of communication protocols and they are complex, customizing hardware for each communication protocol requires a lot of cost and increases the complexity of the system. In addition, with the continuous increase of new protocols, the cost of detection and the complexity of the system will be higher. Summary of the invention

[0005] In view of this, in order to address the above shortcomings, it is necessary to propose an abnormality detection method for a distributed power access unit to reduce the abnormality detection cost and system complexity of the distributed power access unit.

[0006] In a first aspect, the present invention provides an abnormality detection method for a distributed power supply access unit, the abnormality detection method comprising:

[0007] Pre-constructing a simulation model of a photovoltaic inverter; wherein the simulation model includes at least two types of photovoltaic inverters, and each photovoltaic inverter corresponds to a communication protocol;

[0008] Using a distributed power supply access unit to read the simulation model;

[0009] For each photovoltaic inverter, based on the consistency between the photovoltaic inverter data read from the simulation model by the distributed power access unit and the data of the corresponding photovoltaic inverter directly read by the host computer, the inverter data and inverter on / off control of the distributed power access unit are tested;

[0010] According to the inverter data of each type of photovoltaic inverter and the test results of the inverter on / off control of the distributed power access unit, the abnormal state of the distributed power access unit is determined.

[0011] Preferably, the pre-building of the simulation model of the photovoltaic inverter includes:

[0012] Collect historical photovoltaic inverter data of at least two photovoltaic inverters within a period of working time from the photovoltaic network through the master station system;

[0013] Preprocessing the historical photovoltaic inverter data to obtain preprocessed photovoltaic inverter data;

[0014] Performing feature construction on the preprocessed photovoltaic inverter data to obtain a time series feature vector;

[0015] Based on the LSTM model, iterative training is performed using the time series feature vector and the communication protocol to obtain a simulation model including at least two photovoltaic inverters.

[0016] Preferably, the collecting of historical photovoltaic inverter data of at least two photovoltaic inverters within a period of working time from the photovoltaic network through the master station system includes:

[0017] The master station system sends a data acquisition instruction to the concentrator, so that the concentrator sends the data acquisition instruction to the target photovoltaic inverter whose historical photovoltaic inverter data is to be acquired; wherein the historical photovoltaic inverter data includes: voltage, current, power, temperature, and message;

[0018] in,

[0019] After receiving the data acquisition instruction, the target photovoltaic inverter packs the historical photovoltaic inverter data into a data frame and sends the data frame to the master station system; wherein the data frame includes a data header and actual data, and the data header includes a CRC check code obtained by calculating all data;

[0020] The master station system performs data analysis after receiving the data frame and verifies the CRC check code; and obtains corresponding historical photovoltaic inverter data from the data frame that successfully verifies the CRC.

[0021] Preferably, the feature construction of the pre-processed photovoltaic inverter data includes:

[0022] For each type of historical photovoltaic inverter data of each type of photovoltaic inverter, a data identifier of the historical photovoltaic inverter data is combined with a corresponding logical address to obtain a combined feature;

[0023] The combined features of various historical photovoltaic inverters under the same photovoltaic inverter are used to construct a feature vector, and the time series feature vector corresponding to the photovoltaic inverter is obtained.

[0024] Preferably, the iterative training based on the LSTM model using the time series feature vector and the communication protocol includes:

[0025] Initialize input gate, output gate, forget gate and unit state;

[0026] Perform forward propagation calculation by sequentially calculating activation values ​​of an input gate, a forget gate, and an output gate using the time series feature vector;

[0027] The model is trained by calculating the gradient of the loss function to the network parameters based on the back propagation algorithm to obtain a primary simulation model;

[0028] A reliability test is performed on the primary simulation model to obtain a simulation model.

[0029] Preferably, calculating the activation value of the input gate includes calculating using the following formula:

[0030] i t =σ(W xi ·X t +W hi ·h t-1 +W ci ·C t-1 +b i )

[0031] Among them, i t is the activation value of the input gate, σ(·) is the activation function, W xi is the time series feature vector X t The weight matrix to the input gate, W hi is the hidden state h of the previous time step t-1 The weight matrix to the input gate, W ci is the cell state C of the previous time step t-1 The weight matrix to the input gate, b i is the bias term of the input gate;

[0032] Calculating the activation value of the forget gate includes calculating using the following formula:

[0033] f t =σ(W xf ·X t +W hf ·h t-1 +W cf ·C t-1 +b f )

[0034] Among them, f t is the activation value of the forget gate, W xf is the hidden state X of the previous time step t To the weight matrix of the forget gate, W hf is the hidden state h of the previous time step t-1 To the weight matrix of the forget gate, W cf is the cell state C of the previous time step t-1 To the weight matrix of the forget gate, b f is the bias term of the forget gate;

[0035] Calculating the activation value of the output gate includes calculating using the following calculation formula:

[0036] o t =σ(W xo ·X t +W ho ·h t-1 +W co ·C t-1 +b o )

[0037] Among them, t is the activation value of the output gate, W xo is the hidden state X of the previous time step t The weight matrix to the output gate, W ho is the hidden state h of the previous time step t-1 The weight matrix to the output gate, W co is the cell state C of the previous time step t-1 The weight matrix to the output gate, b o is the bias term of the output gate.

[0038] Preferably, the testing of the inverter data of the distributed power access unit based on the consistency of the photovoltaic inverter data read from the simulation model by the distributed power access unit and the data of the corresponding photovoltaic inverter directly read by the host computer comprises:

[0039] Step S301: using the distributed power access unit, reading first inverter data of a first photovoltaic inverter among the at least two types of photovoltaic inverters from the simulation model;

[0040] Step S302: using a host computer to directly read the second inverter data of the first photovoltaic inverter;

[0041] Step S303: determining whether the first inverter data is consistent with the second inverter data;

[0042] Step S304: if the first inverter data is consistent with the second inverter data, then executing step S305, otherwise the distributed power access unit fails to test the data of the first photovoltaic inverter;

[0043] Step S305: using the distributed power access unit to modify the inverter parameters of the first photovoltaic inverter into test inverter data;

[0044] Step S306: directly reading the third inverter data of the first photovoltaic inverter using the host computer;

[0045] Step S307: determining whether the third inverter data is consistent with the test inverter data;

[0046] Step S308: If the third inverter data is consistent with the test inverter data, the distributed power access unit successfully tests the data of the first photovoltaic inverter; otherwise, it fails.

[0047] Preferably, the inverter on / off control of the distributed power access unit is tested based on the consistency of the photovoltaic inverter data read from the simulation model by the distributed power access unit and the data of the corresponding photovoltaic inverter directly read by the host computer, including:

[0048] Step S309: using the distributed power access unit to control the first photovoltaic inverter to start up;

[0049] Step S310: directly reading the first operating state of the first photovoltaic inverter using a host computer;

[0050] Step S311: determining whether the first operating state is a power-on state;

[0051] Step S312: if the first operating state is the power-on state, executing step S313, otherwise the power-on / off control test of the distributed power access unit on the first photovoltaic inverter is unsuccessful;

[0052] Step S313: using the distributed power access unit to control the first photovoltaic inverter to shut down;

[0053] Step S314: directly reading the second operating state of the first photovoltaic inverter using the host computer;

[0054] Step S315: determining whether the second operating state is a shutdown state;

[0055] Step S316: If the second operating state is the shutdown state, the on / off control test of the distributed power supply access unit for the first photovoltaic inverter is successful, otherwise it is unsuccessful.

[0056] Preferably, determining the abnormal state of the distributed power access unit according to the inverter data of each type of photovoltaic inverter and the test results of the inverter on / off control of the distributed power access unit includes:

[0057] For any type of PV inverter, if at least one of the inverter data test and the inverter on / off control test of the distributed power access unit for this type of PV inverter fails, there is an abnormality in the communication of the distributed power access unit for this type of PV inverter.

[0058] In a second aspect, the present invention further provides a computing device, including a memory and a processor, wherein executable code is stored in the memory, and when the processor executes the executable code, any method described in the first aspect is implemented.

[0059] It can be seen from the above technical scheme that when the distributed power access unit is detected for abnormality, the scheme first pre-constructs a simulation model of the photovoltaic inverter, which can simulate at least two types of photovoltaic inverters, and each type of photovoltaic inverter corresponds to a communication protocol, and then uses the distributed power access unit to read the simulation model; further, for each type of photovoltaic inverter, based on the consistency of the inverter data indirectly read by the distributed power access unit and the inverter data directly read by the host computer, the inverter data and inverter switch control of the distributed power access unit are tested, and then the abnormal state of the distributed power access unit can be determined according to the test results. It can be seen that the scheme constructs a simulation model that can simulate photovoltaic inverters of various communication protocol types, and then tests the inverter data and inverter switch control for each type of photovoltaic inverter simulated by the simulation model, so as to determine whether the photovoltaic inverter of the corresponding data type can communicate normally with the distributed power access unit, that is, whether the distributed power access unit is adapted to the communication protocol of the corresponding photovoltaic inverter, and then determine whether the distributed power access unit has a communication abnormality corresponding to a certain communication protocol. In this way, this solution does not need to customize the corresponding hardware equipment for each communication protocol, thereby greatly reducing the cost of abnormal detection of distributed power access units. Moreover, since this solution can be configured and implemented in only one computing device, when there is a new protocol type, it is only necessary to increase the recognition type of the simulation model, and the complexity of the detection system will not be increased. Therefore, this solution can reduce the abnormal detection cost and system complexity of distributed power access units. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 A flow chart of an abnormality detection method for a distributed power supply access unit provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0061] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0062] like Figure 1 As shown, the present invention provides an abnormality detection method for a distributed power supply access unit, and the abnormality detection method may include the following steps:

[0063] Step 1: pre-build a simulation model of a photovoltaic inverter; wherein the simulation model includes at least two types of photovoltaic inverters, and each photovoltaic inverter corresponds to a communication protocol;

[0064] Step 2: Use the distributed power access unit to read the simulation model;

[0065] Step 3: For each type of photovoltaic inverter, based on the consistency of the photovoltaic inverter data read from the simulation model by the distributed power access unit and the data of the corresponding photovoltaic inverter directly read by the host computer, the inverter data and inverter on / off control of the distributed power access unit are tested;

[0066] Step 4: According to the inverter data of the distributed power access unit for each type of photovoltaic inverter and the test results of the inverter on / off control, the abnormal state of the distributed power access unit is determined.

[0067] In this embodiment, a simulation model of photovoltaic inverters including various communication protocol types is constructed, and then the inverter data and inverter switch control are tested for each type of photovoltaic inverter simulated by the simulation model, so as to determine whether the photovoltaic inverter of the corresponding data type can communicate normally with the distributed power access unit, that is, whether the distributed power access unit is adapted to the communication protocol of the corresponding photovoltaic inverter, and then determine whether the distributed power access unit has a communication anomaly of a certain communication protocol. In this way, this solution does not need to customize the corresponding hardware equipment for each communication protocol, so as to greatly reduce the cost of abnormal detection of distributed power access units. Moreover, since this solution can be configured and implemented in only one computing device, when there is a new protocol type, it is only necessary to increase the recognition type of the simulation model, and it will not increase the complexity of the detection system. Therefore, it solves the problem that when testing the adaptability of the distributed power access unit protocol in the past, it is necessary to prepare multiple sets of photovoltaic inverter hardware, resulting in high detection cost and low test efficiency, and this solution ensures the scalability of photovoltaic inverters with new communication protocols.

[0068] For step 1, a simulation model of a photovoltaic inverter is pre-built;

[0069] Since different inverters are adapted or supported by different communication protocols, and the distributed power access unit is a device that converts the communication protocol of the inverter and interacts with the inverter for data, the distributed power access unit must be able to adapt to the communication protocols adapted and compatible by various inverters. When detecting whether the distributed power access unit can adapt to the communication protocols of various inverters, due to the large number of communication protocols, it is necessary to customize the hardware equipment corresponding to the communication protocol, the detection cost is high, and with the continuous increase of new protocols, the detection cost and system complexity will be higher. Based on this, this step considers building a simulation model of the photovoltaic inverter, which should be able to simulate at least two types of photovoltaic inverters, and each photovoltaic inverter corresponds to a communication protocol. In this way, the simulation model is used to simulate photovoltaic inverters of various communication protocols, so that the distributed power access unit can communicate normally with the photovoltaic inverter of the communication protocol, and whether the distributed power access unit is adapted to the communication protocol can be detected, and then it is determined whether the distributed power access unit has a communication protocol adaptation anomaly.

[0070] Specifically, step 1 may include the following steps when pre-building a simulation model of a photovoltaic inverter:

[0071] Step S101: collecting historical photovoltaic inverter data of at least two photovoltaic inverters within a period of working time from the photovoltaic network through the master station system;

[0072] Step S102: preprocessing historical photovoltaic inverter data to obtain preprocessed photovoltaic inverter data;

[0073] Step S103: constructing features for the pre-processed photovoltaic inverter data to obtain a time series feature vector;

[0074] Step S104: Based on the LSTM model, iterative training is performed using the time series feature vector and the communication protocol to obtain a simulation model including at least two photovoltaic inverters.

[0075] For step S101, when collecting historical photovoltaic inverter data of at least two photovoltaic inverters within a period of working time from the photovoltaic network through the master station system, it can be implemented in the following manner:

[0076] The master station system sends a data acquisition instruction to the concentrator, so that the concentrator sends the data acquisition instruction to the target photovoltaic inverter whose historical photovoltaic inverter data is to be acquired; wherein the historical photovoltaic inverter data includes: voltage, current, power, temperature, and message;

[0077] in,

[0078] After receiving the data acquisition instruction, the target photovoltaic inverter packs the historical photovoltaic inverter data into a data frame and sends the data frame to the master station system; wherein the data frame includes a data header and actual data, and the data header includes a CRC check code obtained by calculating all data;

[0079] After receiving the data frame, the master station system performs data analysis and verifies the CRC check code; and obtains the corresponding historical photovoltaic inverter data from the data frame with successful CRC check.

[0080] In this embodiment, the master station sends the information frame containing the request to execute the task to the concentrator, which sends it to the IoT table equipped with the photovoltaic interaction module, and then sends the instruction to the photovoltaic inverter. After receiving the instruction, the photovoltaic inverter reports the key parameters such as voltage, current, power, temperature, and message to the master station. Specifically, after receiving the instruction, the photovoltaic inverter will first collect the working status and performance data of the collector, and then pack the collected data into a data frame. This data frame will include a data header and an actual data part. The data header usually contains metadata, such as the length, type, and CRC check code of the frame. Further in the data header, the inverter will calculate the CRC check code, which is obtained by performing specific mathematical operations on all the data in the data frame to ensure the integrity of the data during transmission. After that, the inverter will send the data frame containing the data header and the actual data to the master station. This process is completed through the communication protocol, which can ensure that the data is transmitted accurately and without error. After receiving the data uploaded by the photovoltaic inverter, the master station first needs to parse the received data frame, which usually involves removing the data header in order to read the actual data information contained. The CRC check code in the data header can check and detect errors that may occur during data transmission, so first check the CRC check code in the data header. If the CRC check fails, it means that the data frame may have been damaged during transmission, and the master station will ignore this data frame. If the CRC check succeeds, the actual data in the data frame can be taken out as historical PV inverter data.

[0081] For step S102, it is considered to pre-process the historical photovoltaic inverter data to obtain pre-processed photovoltaic inverter data. For example, the data under the working state of the photovoltaic inverter is visualized and analyzed, and the data is converted to a unified scale by using data cleaning technology, Z-score standardization and minimum-maximum normalization, etc.

[0082] For step S103, consider constructing features of the pre-processed photovoltaic inverter data to obtain a time series feature vector. This can be specifically achieved in the following manner:

[0083] For each type of historical photovoltaic inverter data of each type of photovoltaic inverter, a data identifier of the historical photovoltaic inverter data is combined with a corresponding logical address to obtain a combined feature;

[0084] The combined features of various historical photovoltaic inverters under the same photovoltaic inverter are used to construct a feature vector, and the time series feature vector corresponding to the photovoltaic inverter is obtained.

[0085] In this embodiment, the pre-processed historical photovoltaic inverter data is enhanced by feature construction and combination, and the OI data identifier of the data packet is combined with the corresponding logical address in the protocol parsing, so as to identify a specific communication protocol.

[0086] For example, the data identifier OI for reading the inverter voltage is 40060200, and the corresponding logical address is 0x60003. After combination, it is the feature {0x60003,40060200}, which indicates the corresponding read inverter voltage instruction under the logical address. Further construct the time series feature vector X_t = [V_t, I_t, P_t, T_t, S_t...], where V_t is the voltage at time point t, I_t is the current at time point t, P_t is the power at time point t, T_t is the temperature at time point t, S_t is the logical address, etc.

[0087] In addition, the time series feature vector can be further standardized.

[0088] For step S104, it is considered to perform iterative training based on the LSTM model using the obtained time series feature vector and the communication protocol to obtain a simulation model that can simulate at least two photovoltaic inverters. Specifically, step S104 can be implemented as follows:

[0089] Initialize input gate, output gate, forget gate and unit state;

[0090] The forward propagation calculation is performed by using the time series feature vector to calculate the activation values ​​of the input gate, forget gate and output gate in sequence;

[0091] The model is trained by calculating the gradient of the loss function to the network parameters based on the back propagation algorithm to obtain a primary simulation model;

[0092] The reliability test is carried out on the primary simulation model to obtain the simulation model.

[0093] In this embodiment, when training a simulation model based on LSTM, the parameters are first initialized, such as initializing the input gate, output gate, forget gate, and unit state, etc., and each parameter in the model is initialized to a constant, and then forward propagation is performed. In the forward propagation, the activation value of the input gate can be calculated by i t =σ(Wxi ·X t +W hi ·h t-1 +W ci ·C t-1 +b i ) is calculated. Among them, i t is the activation value of the input gate, σ(·) is the activation function, W xi is the time series feature vector X t The weight matrix to the input gate, W hi is the hidden state h of the previous time step t-1 The weight matrix to the input gate, W ci is the cell state C of the previous time step t-1 The weight matrix to the input gate, b i is the bias term of the input gate; at each moment, the information input from the input layer will first pass through the input gate, and the switch of the input gate will determine whether there will be information input to the memory cell at this moment.

[0094] When calculating the activation value of the forget gate, f t =σ(W xf ·X t +W hf ·h t-1 +W cf ·C t-1 +b f ) is calculated. Among them, f t is the activation value of the forget gate, W xf is the hidden state X of the previous time step t To the weight matrix of the forget gate, W hf is the hidden state h of the previous time step t-1 To the weight matrix of the forget gate, W cf is the cell state C of the previous time step t-1 To the weight matrix of the forget gate, b f It is the bias term of the forget gate. At each moment, the value in the memory cell will go through a process of being forgotten, which is controlled by this gate. If the clock is punched, the value in the memory cell will be cleared, that is, forgotten.

[0095] When calculating the activation value of the output gate, o t =σ(W xo ·X t +W ho ·h t-1 +W co ·C t-1 +b o ) is calculated. Among them, o t is the activation value of the output gate, W xois the hidden state X of the previous time step t The weight matrix to the output gate, W ho is the hidden state h of the previous time step t-1 The weight matrix to the output gate, W co is the cell state C of the previous time step t-1 The weight matrix to the output gate, b o Is the bias term of the output gate. Whether there is information output from the memory cell at each moment depends on the output gate. The whole process is to first pass through the input gate to see if there is information input, then determine whether the forget gate chooses to forget the information in the memory cell, and finally pass through the output gate to determine whether to output the information at this moment.

[0096] The LSTM network is similar to the traditional RNN in the training process, relying on the back-propagation algorithm to calculate the gradient of the loss function to the network parameters. However, the gate unit in the LSTM increases the complexity of the gradient calculation. To deal with this problem, the weighted back-propagation method can be used, which adjusts the weight of the gate unit gradient to enhance its influence on the overall gradient. The specific operation is to multiply the output of the gate unit by its input to generate a weight factor, and then apply this weight factor to the gradient of the gate unit, which can effectively simplify the calculation process of the gradient in the LSTM network.

[0097] Furthermore, the validation set can be used to detect the trained LSTM model, connect a new photovoltaic inverter whose data is not used in the trained LSTM model, test the reliability of the trained LSTM model, and continuously optimize and improve the model.

[0098] For step 2, the simulation model is read using the distributed power access unit;

[0099] After completing the training of the simulation model, consider using the transfer learning method, based on the trained LSTM model, combined with software to achieve dynamic switching of the photovoltaic inverter communication protocol to simulate the actual installation scenario of the distributed power access unit.

[0100] For step 3, for each type of photovoltaic inverter, based on the consistency of the photovoltaic inverter data read from the simulation model by the distributed power access unit and the data of the corresponding photovoltaic inverter directly read by the host computer, the inverter data and inverter on / off control of the distributed power access unit are tested;

[0101] In this step, the interaction of the distributed power access unit with respect to the inverter data and the control of the inverter on / off are tested respectively.

[0102] When testing the inverter data of the distributed power access unit, the following steps may be included:

[0103] Step S301: using a distributed power source access unit, reading first inverter data of a first photovoltaic inverter of at least two types of photovoltaic inverters from a simulation model;

[0104] Step S302: directly reading the second inverter data of the first photovoltaic inverter using the host computer;

[0105] Step S303: determining whether the first inverter data is consistent with the second inverter data;

[0106] Step S304: if the data of the first inverter is consistent with the data of the second inverter, then execute step S305, otherwise the data test of the distributed power access unit on the first photovoltaic inverter is unsuccessful;

[0107] Step S305: using the distributed power access unit to modify the inverter parameters of the first photovoltaic inverter to test inverter data;

[0108] Step S306: directly reading the third inverter data of the first photovoltaic inverter using the host computer;

[0109] Step S307: determining whether the third inverter data is consistent with the test inverter data;

[0110] Step S308: If the third inverter data is consistent with the test inverter data, the distributed power access unit successfully tests the data of the first photovoltaic inverter; otherwise, it fails.

[0111] In this embodiment, the simulation model can simulate the current photovoltaic inverter. After the simulated photovoltaic inverter is connected to the distributed power access unit, the inverter data of the photovoltaic inverter is read through the distributed power access unit, and the inverter data of the photovoltaic inverter can be directly read through the host computer. In this way, if the inverter data read through the distributed power access unit is consistent with the inverter data directly read through the host computer, it means that the current photovoltaic inverter can perform normal data interaction with the distributed power access unit; if it is inconsistent, it means that the distributed power access unit cannot adapt to the communication protocol of the current photovoltaic inverter, and the two cannot perform normal data interaction, and then the inverter data test of the distributed power access unit fails.

[0112] When the inverter data read through the distributed power access unit is consistent with the inverter data directly read through the host computer, the inverter parameters of the photovoltaic inverter are further modified to the test inverter data using the distributed power access unit, and then the third inverter data of the photovoltaic inverter is directly read again using the host computer. If the read third inverter data is consistent with the modified test inverter data, it means that the photovoltaic inverter parameters can be adjusted through the distributed power access unit and the test is successful; if the read third inverter data is inconsistent with the modified test inverter data, it means that the photovoltaic inverter parameters cannot be adjusted through the distributed power access unit, there is a problem of communication protocol compatibility between the two, and the test is unsuccessful.

[0113] When testing the inverter on / off control of the distributed power access unit, the following steps may be included:

[0114] Step S309: using the distributed power supply access unit to control the first photovoltaic inverter to start up;

[0115] Step S310: directly reading the first operating state of the first photovoltaic inverter using the host computer;

[0116] Step S311: determining whether the first operating state is a power-on state;

[0117] Step S312: if the first operating state is the power-on state, executing step S313, otherwise the power-on / off control test of the distributed power access unit for the first photovoltaic inverter is unsuccessful;

[0118] Step S313: using the distributed power supply access unit to control the first photovoltaic inverter to shut down;

[0119] Step S314: directly reading the second operating state of the first photovoltaic inverter using the host computer;

[0120] Step S315: determining whether the second operating state is a shutdown state;

[0121] Step S316: If the second operating state is the shutdown state, the on / off control test of the distributed power supply access unit for the first photovoltaic inverter is successful, otherwise it is unsuccessful.

[0122] In this embodiment, after testing the inverter data of the distributed power access unit, the distributed power access unit can also test the power on and off control of the inverter. Specifically, the photovoltaic inverter can be turned on by the distributed power access unit, and then the operating status of the photovoltaic inverter can be directly read by the host computer. By judging whether the operating status read by the host computer is the power-on state, it can be determined whether the distributed power access unit can normally control the power on of the current photovoltaic inverter. When the power-on test is successful, the shutdown is further tested.

[0123] Of course, it should be pointed out that the inverter data test and the inverter power on / off test can not only be tested simultaneously to determine the abnormal state of the distributed power access unit, but in some embodiments, only one of the two can be tested.

[0124] For step 104, the abnormal state of the distributed power access unit is determined according to the inverter data of each type of photovoltaic inverter and the test results of the inverter on / off control of the distributed power access unit;

[0125] In this step, after completing the tests of the inverter data and inverter on / off control of each type of photovoltaic inverter, consider determining the abnormal state of the distributed power access unit based on the test results. Specifically, for any type of photovoltaic inverter, if at least one of the inverter data test and inverter on / off control test of the distributed power access unit for the photovoltaic inverter of this type fails, the communication of the distributed power access unit for the photovoltaic inverter of this type is abnormal.

[0126] In this embodiment, if the inverter data test or inverter power on / off control test of a certain type of photovoltaic inverter is unsuccessful, it means that the distributed power access unit cannot effectively adapt to the communication protocol corresponding to the photovoltaic inverter, and further indicates that the distributed power access unit has an abnormality that cannot adapt to the communication protocol.

[0127] The present specification also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed in a computer, the computer is caused to execute a method in any one of the embodiments in the specification.

[0128] The present specification also provides a computing device, including a memory and a processor, wherein executable codes are stored in the memory, and when the processor executes the executable codes, a method in any embodiment of the present specification is implemented.

[0129] In summary, the method for detecting anomalies of a distributed power access unit provided by the present invention has at least the following beneficial effects:

[0130] 1. By using the LSTM deep learning model, this solution can effectively process sequence data such as voltage, current, and power and capture time dependencies, thereby more accurately simulating various types of photovoltaic inverters. Feature enhancement operations and data preprocessing operations help improve the quality of data, thereby improving the model's prediction ability and detection efficiency.

[0131] 2. Generally, adaptive anomaly detection of distributed power access units is difficult to complete by preparing multiple types of photovoltaic inverters and other hardware equipment, resulting in inaccurate anomaly detection results. This solution obtains a simulation model based on LSTM deep learning training, and uses various types of simulated photovoltaic inverters to test distributed power access units. It uses a combination of software and hardware to reduce hardware costs and testing time, and improve testing efficiency.

[0132] 3. This solution is capable of transfer learning, which improves the simulation model’s adaptation speed and accuracy to new protocols and ensures the scalability of the system.

[0133] 4. Compared with the LSTM algorithm commonly used in fault diagnosis, the GRU in this solution reduces two gating signals, reduces the computational complexity, and improves the real-time performance of model anomaly detection.

[0134] 5. This solution can archive and analyze detected abnormal data, generate statistical reports and improvement suggestions, and help continuously optimize the system and improve overall performance and reliability.

[0135] Since the device embodiments provided by the present invention are based on the same inventive concept as the method embodiments of this specification, the specific contents can be found in the description of the method embodiments of this specification and will not be repeated here.

[0136] The modules or units in the device of the embodiment of the present invention can be combined, divided and deleted according to actual needs. The above disclosure is only the preferred embodiment of the present invention, and of course it cannot be used to limit the scope of the rights of the present invention. Those skilled in the art can understand that all or part of the processes of the above embodiment are implemented, and the equivalent changes made according to the claims of the present invention still fall within the scope of the invention.

Claims

1. A method for detecting anomalies of a distributed power supply access unit, characterized in that: The anomaly detection method includes: Pre-constructing a simulation model of a photovoltaic inverter; wherein the simulation model includes at least two types of photovoltaic inverters, and each photovoltaic inverter corresponds to a communication protocol; Using a distributed power supply access unit to read the simulation model; For each photovoltaic inverter, based on the consistency between the photovoltaic inverter data read from the simulation model by the distributed power access unit and the data of the corresponding photovoltaic inverter directly read by the host computer, the inverter data and inverter on / off control of the distributed power access unit are tested; According to the inverter data of each type of photovoltaic inverter and the test results of the inverter on / off control of the distributed power access unit, the abnormal state of the distributed power access unit is determined.

2. The abnormality detection method of the distributed power access unit according to claim 1, characterized in that: The pre-built simulation model of the photovoltaic inverter includes: Collect historical photovoltaic inverter data of at least two photovoltaic inverters within a period of working time from the photovoltaic network through the master station system; Preprocessing the historical photovoltaic inverter data to obtain preprocessed photovoltaic inverter data; Performing feature construction on the preprocessed photovoltaic inverter data to obtain a time series feature vector; Based on the LSTM model, iterative training is performed using the time series feature vector and the communication protocol to obtain a simulation model including at least two photovoltaic inverters.

3. The abnormality detection method of the distributed power access unit according to claim 2, characterized in that: The collecting of historical photovoltaic inverter data of at least two photovoltaic inverters within a period of working time from the photovoltaic network through the master station system includes: The master station system sends a data acquisition instruction to the concentrator, so that the concentrator sends the data acquisition instruction to the target photovoltaic inverter whose historical photovoltaic inverter data is to be acquired; wherein the historical photovoltaic inverter data includes: voltage, current, power, temperature, and message; in, After receiving the data acquisition instruction, the target photovoltaic inverter packs the historical photovoltaic inverter data into a data frame and sends the data frame to the master station system; wherein the data frame includes a data header and actual data, and the data header includes a CRC check code obtained by calculating all data; The master station system performs data analysis after receiving the data frame and verifies the CRC check code; and obtains corresponding historical photovoltaic inverter data from the data frame that successfully verifies the CRC.

4. The abnormality detection method of the distributed power access unit according to claim 2, characterized in that: The feature construction of the pre-processed photovoltaic inverter data includes: For each type of historical photovoltaic inverter data of each type of photovoltaic inverter, a data identifier of the historical photovoltaic inverter data is combined with a corresponding logical address to obtain a combined feature; The combined features of various historical photovoltaic inverters under the same photovoltaic inverter are used to construct a feature vector, and the time series feature vector corresponding to the photovoltaic inverter is obtained.

5. The abnormality detection method of a distributed power supply access unit according to claim 2, characterized in that: The iterative training based on the LSTM model using the time series feature vector and the communication protocol includes: Initialize input gate, output gate, forget gate and unit state; Perform forward propagation calculation by sequentially calculating activation values ​​of an input gate, a forget gate, and an output gate using the time series feature vector; The model is trained by calculating the gradient of the loss function to the network parameters based on the back propagation algorithm to obtain a primary simulation model; A reliability test is performed on the primary simulation model to obtain a simulation model.

6. The abnormality detection method for a distributed power supply access unit according to claim 5, characterized in that: Calculating the activation value of the input gate includes calculating using the following calculation formula: i t =σ(W xi ·X t +W hi ·h t-1 +W ci ·C t-1 +b i ) Among them, i t is the activation value of the input gate, σ(·) is the activation function, W xi is the time series feature vector X t The weight matrix to the input gate, W hi is the hidden state h of the previous time step t-1 The weight matrix to the input gate, W ci is the cell state C of the previous time step t-1 The weight matrix to the input gate, b i is the bias term of the input gate; Calculating the activation value of the forget gate includes calculating using the following formula: f t =σ(W xf ·X t +W hf ·h t-1 +W cf ·C t-1 +b f ) Among them, f t is the activation value of the forget gate, W xf is the hidden state X of the previous time step t To the weight matrix of the forget gate, W hf is the hidden state h of the previous time step t-1 To the weight matrix of the forget gate, W cf is the cell state C of the previous time step t-1 To the weight matrix of the forget gate, b f is the bias term of the forget gate; Calculating the activation value of the output gate includes calculating using the following calculation formula: o t =σ(W xo ·X t +W ho ·h t-1 +W co ·C t-1 +b o ) Among them, t is the activation value of the output gate, W xo is the hidden state X of the previous time step t The weight matrix to the output gate, W ho is the hidden state h of the previous time step t-1 The weight matrix to the output gate, W co is the cell state C of the previous time step t-1 The weight matrix to the output gate, b o is the bias term of the output gate.

7. The abnormality detection method of a distributed power supply access unit according to claim 1, characterized in that: The testing of the inverter data of the distributed power access unit based on the consistency of the photovoltaic inverter data read from the simulation model by the distributed power access unit and the data of the corresponding photovoltaic inverter directly read by the host computer includes: Step S301: using the distributed power access unit, reading first inverter data of a first photovoltaic inverter among the at least two types of photovoltaic inverters from the simulation model; Step S302: using a host computer to directly read the second inverter data of the first photovoltaic inverter; Step S303: determining whether the first inverter data is consistent with the second inverter data; Step S304: if the first inverter data is consistent with the second inverter data, then executing step S305, otherwise the distributed power access unit fails to test the data of the first photovoltaic inverter; Step S305: using the distributed power access unit to modify the inverter parameters of the first photovoltaic inverter into test inverter data; Step S306: directly reading the third inverter data of the first photovoltaic inverter using the host computer; Step S307: determining whether the third inverter data is consistent with the test inverter data; Step S308: If the third inverter data is consistent with the test inverter data, the distributed power access unit successfully tests the data of the first photovoltaic inverter; otherwise, it fails.

8. The method for detecting abnormality of a distributed power access unit according to claim 7, characterized in that: The inverter on / off control of the distributed power access unit is tested based on the consistency of the photovoltaic inverter data read from the simulation model by the distributed power access unit and the data of the corresponding photovoltaic inverter directly read by the host computer, including: Step S309: using the distributed power access unit to control the first photovoltaic inverter to start up; Step S310: directly reading the first operating state of the first photovoltaic inverter using a host computer; Step S311: determining whether the first operating state is a power-on state; Step S312: if the first operating state is the power-on state, executing step S313, otherwise the power-on / off control test of the distributed power access unit on the first photovoltaic inverter is unsuccessful; Step S313: using the distributed power access unit to control the first photovoltaic inverter to shut down; Step S314: directly reading the second operating state of the first photovoltaic inverter using the host computer; Step S315: determining whether the second operating state is a shutdown state; Step S316: If the second operating state is the shutdown state, the on / off control test of the distributed power supply access unit for the first photovoltaic inverter is successful, otherwise it is unsuccessful.

9. The abnormality detection method of the distributed power access unit according to claim 8, characterized in that: The determining of the abnormal state of the distributed power access unit according to the inverter data of each type of photovoltaic inverter and the test results of the inverter on / off control of the distributed power access unit comprises: For any type of PV inverter, if at least one of the inverter data test and the inverter on / off control test of the distributed power access unit for this type of PV inverter fails, there is an abnormality in the communication of the distributed power access unit for this type of PV inverter.

10. A computing device, comprising a memory and a processor, wherein the memory stores executable codes, and when the processor executes the executable codes, the method according to any one of claims 1 to 9 is implemented.

Images