Method and device for automatically testing transformer substation fault and electronic equipment

Abstract

The invention provides an automatic testing method and device for transformer substation faults and electronic equipment. The method comprises the steps that test parameters are acquired, and the test parameters are parameters for performing relay protection test on the transformer substation; the test parameters are analyzed through a lightweight gradient lifting machine learning model, a test result is obtained, each group of data in the multiple groups of data comprises historical test parameters and historical test results, and the test results are obtained by testing the transformer substation; and determining the fault type of the substation according to the test result. Therefore, the test parameters are intelligently analyzed through the lightweight gradient lifting machine learning model, the fault type of the transformer substation can be automatically distinguished, misjudgment is reduced, and the problem of misjudgment in the automatic fault test technology of the transformer substation in the prior art is solved.

Description

Technical Field

[0001] The present application relates to the field of substation fault detection, and in particular, to an automatic testing method, device, computer-readable storage medium and electronic device for substation faults. Background Art

[0002] With the promotion and construction of smart grids, the traditional way of transmitting electrical signals using cables as the medium has been replaced by optical fiber digitalization. Information is no longer a single point-to-point communication method, but can be transmitted by building a new communication network. At the same time, the correlation between a single device and other devices in a smart substation is also closer, and some new protection devices need to be judged based on the output information of multiple devices. How to verify these complex information organization relationships and comprehensive protection and control functions is a problem that must be solved in the inspection work before the commissioning of smart substations. In order to meet the needs of the development of smart substations, different smart substation testing methods and tools have been proposed and developed in the prior art. At present, most of the testing tools are mainly equipment testers.

[0003] Applying automated testing technology in smart substations to automatically complete system testing tasks can theoretically save testing time, but there are many limitations in actual situations. For example, if the test results are incorrect, it will take a lot of energy and time to analyze and study the reasons. It is understood that the current tester cannot analyze the reasons for the operation of the equipment, but can only determine the results of the test. There are great limitations, and the equipment analysis is not intelligent enough. In addition, for example, the differential I section of the line protection overlaps with other protection logics. During the test, there will be logic mismatches such as protection pressure plates. In the end, the test logic judgment of other protections may cover the test logic judgment of the current protection, and finally misjudge the successful action, which brings inconvenience to the operation and maintenance work and there are cases of test misjudgment. From the above problems, it can be seen that the automatic testing technology is not intelligent enough, and the analysis of the results needs to be further optimized. Summary of the invention

[0004] The main purpose of the present application is to provide a method, device, computer-readable storage medium and electronic device for automatic testing of substation faults, so as to at least solve the problem of misjudgment in the prior art automatic testing technology for substation faults.

[0005] In order to achieve the above-mentioned purpose, according to one aspect of the present application, there is provided an automatic testing method for substation faults, comprising: obtaining test parameters, wherein the test parameters are parameters for performing relay protection tests on the substation; analyzing the test parameters through a lightweight gradient boosting machine learning model to obtain test results, wherein the lightweight gradient boosting machine learning model is obtained through machine learning training based on multiple groups of data, each group of the multiple groups of data includes: historical test parameters and historical test results, and the test results are the results obtained by testing the substation; determining the type of fault occurring in the substation according to the test results.

[0006] Optionally, before obtaining the test parameters, the method also includes: obtaining the historical test parameters and the historical test results, constructing an input vector based on the historical test parameters, and constructing an output vector based on the historical test results; obtaining an initial lightweight gradient boosting machine learning model, training the initial lightweight gradient boosting machine learning model through the input vector and the output vector to obtain optimal model parameters, configuring the model parameters of the initial lightweight gradient boosting machine learning model to the optimal model parameters, and obtaining the lightweight gradient boosting machine learning model.

[0007] Optionally, obtaining the historical test parameters and the historical test results includes: obtaining at least the soft pressure plate parameters, control word parameters, zero-sequence overcurrent protection parameters and relay protection setting parameters of the substation to obtain the historical test parameters; obtaining the test results of the historical test parameters to obtain the historical test results.

[0008] Optionally, an input vector is constructed according to the historical test parameters, and an output vector is constructed according to the historical test results, including: constructing the input vector according to different setting values ​​of the historical test parameters, wherein the setting values ​​include 0 and 1, 0 represents an open state, and 1 represents a closed state; assigning a fault code to the test results representing the faults in the historical test results to construct the output vector.

[0009] Optionally, the initial lightweight gradient boosting machine learning model is trained by the input vector and the output vector to obtain the optimal model parameters, including: determining multiple groups of initial model parameters, wherein each group of the initial model parameters includes multiple model parameters, one of the model parameters changes, and the remaining model parameters remain unchanged; under each group of the initial model parameters, the initial lightweight gradient boosting machine learning model is trained by the input vector and the output vector to obtain the accuracy of the training results; and determining the group of initial model parameters with the highest accuracy as the optimal model parameters.

[0010] Optionally, determining the fault type occurring in the substation according to the test result includes: obtaining a one-to-one mapping relationship between fault type and fault code; and determining the fault type corresponding to the fault code in the test result according to the one-to-one mapping relationship between the fault type and fault code.

[0011] Optionally, after obtaining the test results, the method further includes: calculating the prediction accuracy for each fault type, and adjusting the optimal model parameters according to the prediction accuracy, wherein the prediction accuracy is the ratio of the number of correctly predicted data to the total number of predicted data.

[0012] According to another aspect of the present application, an automatic testing device for substation faults is provided, including: an acquisition unit, used to acquire test parameters, wherein the test parameters are parameters for relay protection testing of the substation; an analysis unit, used to analyze the test parameters through a lightweight gradient boosting machine learning model to obtain test results, wherein the lightweight gradient boosting machine learning model is obtained through machine learning training based on multiple groups of data, each group of the multiple groups of data includes: historical test parameters and historical test results, and the test results are results obtained by testing the substation; a determination unit, used to determine the type of fault occurring in the substation based on the test results.

[0013] According to another aspect of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute any one of the automatic testing methods for substation faults.

[0014] According to another aspect of the present application, an electronic device is provided, comprising: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include an automatic testing method for executing any one of the substation faults.

[0015] Apply the technical solution of the present application to obtain test parameters, wherein the test parameters are parameters for relay protection testing of substations; analyze the test parameters through a lightweight gradient boosting machine learning model to obtain test results, wherein the lightweight gradient boosting machine learning model is obtained through machine learning training based on multiple groups of data, each of the multiple groups of data includes: historical test parameters and historical test results, and the test results are the results obtained by testing the substation; determine the type of fault that occurred in the substation based on the test results. In this way, by performing intelligent analysis of the test parameters through the above-mentioned lightweight gradient boosting machine learning model, the type of fault that occurred in the substation can be automatically identified, reducing the occurrence of misjudgment, and solving the problem of misjudgment in the automatic fault testing technology of substations in the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The drawings constituting part of the present application are used to provide a further understanding of the present application. The exemplary embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0017] Figure 1 A hardware structure block diagram of a mobile terminal for performing an automatic testing method for a substation fault provided in an embodiment of the present application is shown;

[0018] Figure 2 A schematic diagram of a flow chart of an automatic testing method for a substation fault provided by an embodiment of the present application is shown;

[0019] Figure 3 A structural block diagram of an automatic testing device for substation faults provided in an embodiment of the present application is shown.

[0020] The above drawings include the following reference numerals:

[0021] 102, processor; 104, memory; 106, transmission device; 108, input and output devices. DETAILED DESCRIPTION

[0022] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0023] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0024] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present application described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0025] For the convenience of description, some nouns or terms involved in the embodiments of the present application are explained below:

[0026] The Lightweight Gradient Boosting Machine Learning Model (LGBM) is a machine learning algorithm based on gradient boosting. Compared with traditional gradient boosting machines (such as XGBoost and LightGBM), the lightweight gradient boosting model is more efficient and lightweight, and is suitable for running in resource-constrained environments. LGBM improves the training speed and prediction performance of the model by optimizing the algorithm and reducing the computational complexity.

[0027] As introduced in the background technology, the tester in the prior art cannot analyze the cause of the operation of the equipment and can only determine the result of the test, which has great limitations. The equipment analysis is not intelligent enough. In order to solve the problem of misjudgment in the automatic fault testing technology of substations in the prior art, the embodiments of the present application provide a method, device, computer-readable storage medium and electronic device for automatic testing of substation faults.

[0028] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0029] The method embodiments provided in the embodiments of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Taking running on a mobile terminal as an example, Figure 1 FIG. 1 is a hardware structure block diagram of a mobile terminal of an automatic testing method for a substation fault according to an embodiment of the present invention. Figure 1 As shown, the mobile terminal may include one or more ( Figure 1 Only one is shown in the figure) a processor 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data, wherein the mobile terminal may also include a transmission device 106 and an input / output device 108 for communication functions. It can be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the mobile terminal. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations are shown.

[0030] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the automatic test method for substation faults in the embodiment of the present invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, the above method is implemented. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories can be connected to the mobile terminal via a network. Examples of the above network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof. The transmission device 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the mobile terminal. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0031] In this embodiment, a method for automatically testing substation faults running on a mobile terminal, a computer terminal or a similar computing device is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0032] Figure 2 FIG. 1 is a flow chart of an automatic testing method for a substation fault according to an embodiment of the present application. Figure 2 As shown, the method comprises the following steps:

[0033] Step S201, obtaining test parameters, wherein the test parameters are parameters for performing relay protection tests on the substation;

[0034] Specifically, the relay protection tester can simulate the intelligent terminal and the merging unit, and the general template file of the tester contains template information, protection test items and other parameter categories. According to the test results of the relay protection tester in the automatic test process and analysis of parameters that can be used, such as protection test phase setting, intelligent terminal switch status, merging unit current frequency, merging unit voltage frequency, merging unit voltage amplitude, etc., the system needs to obtain various parameters of the substation when conducting relay protection testing. These parameters may include electrical parameters such as voltage, current, temperature, frequency, as well as relay action time, protection setting, etc. These parameters are the basic data for fault diagnosis. By accurately obtaining these parameters, reliable input data can be provided for subsequent fault identification.

[0035] Step S202, analyzing the test parameters through a lightweight gradient boosting machine learning model to obtain a test result, wherein the lightweight gradient boosting machine learning model is obtained through machine learning training based on multiple sets of data, each set of data in the multiple sets of data includes: historical test parameters and historical test results, and the test result is a result obtained by testing the substation;

[0036] Specifically, the model is trained based on a large amount of historical data, each set of which includes historical test parameters and corresponding historical test results. By learning from these data, the model can identify the complex relationship between test parameters and fault types, thereby generating prediction results for the current test.

[0037] Step S203: determine the type of fault that occurs in the substation according to the test result.

[0038] Specifically, the system determines the type of fault currently occurring in the substation based on the prediction results of the model. By analyzing the test results, the system can identify the specific fault type, such as soft pressure plate configuration error, control word configuration error, etc., and make corresponding responses or warnings.

[0039] Through this embodiment, firstly, the test parameters are obtained, wherein the test parameters are the parameters for the relay protection test of the substation; the test parameters are analyzed by the lightweight gradient boosting machine learning model to obtain the test results, wherein the lightweight gradient boosting machine learning model is obtained by machine learning training based on multiple groups of data, and each group of data in the multiple groups of data includes: historical test parameters and historical test results, and the test results are the results obtained by testing the substation; the type of fault that occurred in the substation is determined according to the test results. In this way, by intelligently analyzing the test parameters through the above-mentioned lightweight gradient boosting machine learning model, the type of fault that occurred in the substation can be automatically identified, the occurrence of misjudgment can be reduced, the accuracy and efficiency of fault diagnosis can be improved, and the problem of misjudgment in the automatic fault testing technology of the substation in the prior art is solved.

[0040] In the specific implementation process, the above step S201 can be implemented by the following steps:

[0041] Step 301, obtaining historical test parameters and historical test results, constructing an input vector according to the historical test parameters, and constructing an output vector according to the historical test results;

[0042] Specifically, the system first obtains a large number of historical test parameters and corresponding historical test results. Then, it constructs input vectors (historical test parameters) and output vectors (historical test results) based on these historical data. The input vector is used as input data for model training, and the output vector is used as a training target to help the model learn the relationship between parameters and fault types.

[0043] Step 302, obtain an initial lightweight gradient boosting machine learning model, train the initial lightweight gradient boosting machine learning model through input vectors and output vectors to obtain optimal model parameters, configure the model parameters of the initial lightweight gradient boosting machine learning model as the optimal model parameters, and obtain a lightweight gradient boosting machine learning model.

[0044] Gradient boosting is a commonly used machine learning method, and there are many methods of gradient boosting. Gradient boosting decision tree (GBDT) is a common gradient boosting method. As the basis function of the GBDT algorithm, the decision tree adds sub-models to synthesize a strong regression tree in order to continuously reduce the loss function in a one-time iterative calculation. The system uses a pre-trained lightweight gradient boosting machine learning model to analyze the obtained test parameters.

[0045] The training and testing model framework based on the Light GBM algorithm is divided into four parts: sample data set segmentation, parameter tuning, model training, and data testing. Data set segmentation is to randomly select 70% of them as training data and the remaining 30% as test data, and use cross-validation to score the model and draw a learning curve. According to the requirements of the automatic test results of the automatic testing technology, the most important parameters are selected to determine the learning rate and the number of leaf nodes, and the default values ​​are selected for other parameters. By analyzing the parameters of the algorithm model, adjusting the most important number of leaf nodes num_leaves, find the optimal parameters used by the training data set on this model architecture. Then use the determined optimal parameters to continuously train the model, generate and save the optimal model for judging the test data results. Finally, the saved optimal model is tested with data and applied to the analysis and judgment of the automatic test results.

[0046] Specifically, the system obtains the initial lightweight gradient boosting machine learning model and trains the model through input vectors and output vectors. During the training process, the system adjusts the model parameters to achieve the best prediction effect. Finally, the model parameters are configured as the optimal model parameters, and a lightweight gradient boosting machine learning model that can be used for actual testing is obtained. It can be seen that this method can train a machine learning model that can accurately identify the type of substation faults, providing reliable support for subsequent fault diagnosis.

[0047] In order to obtain historical test parameters and historical test results, the above step S301 of the present application can be implemented by the following steps:

[0048] Step 401, at least obtain the soft pressure plate parameters, control word parameters, zero-sequence overcurrent protection parameters and relay protection setting parameters of the substation to obtain historical test parameters;

[0049] Specifically, the soft pressure plate is a control device in the substation, which is used to control the switching of the protection device. Obtaining the soft pressure plate parameters can understand the working status of the protection device; the control word parameters are key parameters in the substation control system, which are used to indicate various operations and states. Obtaining these parameters helps to understand the overall operating status of the system; zero-sequence overcurrent protection is a protection mechanism in the substation, which is used to detect ground faults. Obtaining these parameters can understand the performance of ground protection; the relay protection setting parameter is the action threshold of the relay under various fault conditions. Obtaining these parameters can understand the protection capability of the relay. By obtaining these key historical test parameters, the system provides the necessary data basis for subsequent model training. These parameters can reflect the operating status and protection performance of the substation, and are an important basis for fault diagnosis.

[0050] Step 402, obtaining the test results of the historical test parameters to obtain the historical test results.

[0051] Specifically, the method needs to obtain the test results corresponding to the above historical test parameters. These test results are usually fault records of substations in different states, including fault type, fault time, fault location and other information. Obtaining these test results can provide accurate output data for model training. It can be seen that by obtaining the test results of historical test parameters, the system can construct a complete training data set (input vector and output vector), providing a reliable basis for the training of lightweight gradient boosting machine learning models.

[0052] In order to construct the input vector and the output vector, the above step S301 of the present application can be implemented by the following steps:

[0053] Step 501, constructing an input vector according to different setting values ​​of historical test parameters, wherein the setting value includes 0 and 1, 0 represents an open state, and 1 represents a closed state;

[0054] Specifically, when a parameter is 0, it indicates that the corresponding device or protection mechanism is in the disconnected state; when a parameter is 1, it indicates that the corresponding device or protection mechanism is in the closed state or working state. By combining these setting values, the system can construct an input vector containing a variety of state information for model training. Different combinations of setting values ​​can reflect the operation status of the substation in various states and provide rich data support for fault diagnosis.

[0055] Step 502: assign fault codes to the test results representing faults in the historical test results, and construct an output vector.

[0056] Specifically, the fault code can be a number or a character to identify different fault types, for example: 1 indicates a soft pressure plate configuration error, 2 a control word configuration error, 3 a fixed value configuration error, 4 indicates a large current\voltage error, etc. By combining these fault codes, the system can construct an output vector for the training target of the model. It can be seen that this method can convert the fault types in the test results into standard fault codes, which is convenient for model learning and recognition. These fault codes can accurately reflect different fault types and provide clear targets for fault diagnosis.

[0057] Through the above steps, the system can obtain the historical test parameters and test results of the substation in detail and convert them into input vectors and output vectors suitable for model training. This process provides the necessary data support for the training of the lightweight gradient boosting machine learning model, ensuring that the model can accurately identify the fault type of the substation and reduce the occurrence of misjudgment.

[0058] In some embodiments, the above step S302 can be implemented by the following steps:

[0059] Step S601, determining multiple groups of initial model parameters, wherein each group of initial model parameters includes multiple model parameters, one of which is changed, while the remaining model parameters remain unchanged;

[0060] Specifically, multiple groups of initial model parameters are first determined. Each group of initial model parameters includes multiple model parameters, one of which is changed, while the other model parameters remain unchanged. For example, if there are three model parameters, A, B, and C, then one of A, B, and C can be changed respectively, while the other two parameters are kept unchanged, forming three groups of initial model parameters. This method can explore the combination of multiple model parameters, find the possible optimal parameter combination, and provide multiple starting points for subsequent model training.

[0061] Step S602, under each set of initial model parameters, train the initial lightweight gradient boosting machine learning model through the input vector and the output vector to obtain the accuracy of the training result;

[0062] Specifically, under each set of initial model parameters, the system uses the input vector (historical test parameters) and the output vector (historical test results) to train the initial lightweight gradient boosting machine learning model. After the training is completed, the system calculates the accuracy of the training results, that is, the model's prediction accuracy for the historical test data. Through this method, the system can evaluate the performance of the model under different combinations of model parameters, thereby screening out the parameter combination with the highest accuracy.

[0063] Step S603: determine a set of initial model parameters with the highest accuracy as the optimal model parameters.

[0064] Specifically, among all the initial model parameter combinations, the system determines the set of initial model parameters with the highest accuracy as the optimal model parameters, and assigns them to the initial lightweight gradient boosting machine learning model, thereby obtaining a lightweight gradient boosting machine learning model that can be used for actual testing. Through this method, the system can ensure the optimization of model parameters and improve the prediction accuracy of the model in practical applications.

[0065] In order to determine the type of fault that occurs in the substation, the above step S203 of the present application can be implemented by the following steps:

[0066] Step S701, obtaining a one-to-one mapping relationship between fault types and fault codes;

[0067] Specifically, first obtain a one-to-one mapping relationship between fault types and fault codes. For example, a short circuit fault corresponds to fault code 1, a ground fault corresponds to fault code 2, an overload fault corresponds to fault code 3, an undervoltage fault corresponds to fault code 4, etc. This step can establish a corresponding relationship between fault types and fault codes, providing a basis for subsequent fault type identification.

[0068] Step S702: determine the fault type corresponding to the fault code in the test result according to the one-to-one mapping relationship between the fault type and the fault code.

[0069] Specifically, the system parses the fault code in the model prediction result according to the mapping relationship between the acquired fault type and fault code to determine the corresponding fault type. For example, if the fault code in the model prediction result is 2, the system will determine that a ground fault has occurred in the substation. This method system can accurately identify the fault type of the substation and provide clear guidance for subsequent fault handling.

[0070] In some embodiments, the above step S203 further includes the following steps: calculating the prediction accuracy for each fault type, and adjusting the optimal model parameters according to the prediction accuracy, wherein the prediction accuracy is the ratio of the number of correctly predicted data to the total number of predicted data.

[0071] The comparison of the accuracy of prediction errors of different fault types analyzed by the training model under multiple protection logics is shown in Table 1.

[0072] Table 1. Correctness under multiple logics

[0073]

[0074] Specifically, the prediction accuracy is calculated for each fault type. The prediction accuracy is the ratio of the number of correctly predicted data to the total number of predicted data. According to the calculated prediction accuracy of each fault type, the system analyzes the fault type with low prediction accuracy, finds out the model parameters that affect the prediction accuracy, and adjusts them. The system can improve the prediction accuracy by increasing training data, adjusting the model structure, or optimizing model parameters. Analysis of the above table shows that when the learning rate is 0.1, the L1 regularization penalty term is 0.1, the L2 regularization penalty term is 0.2, the maximum depth of the tree is 4, and the number of leaf nodes is about 30, the accuracy of the simulation training data is the highest, the determined model parameters are optimal, and the optimal training model is obtained by training with the optimal parameters. The accuracy of each fault type is basically higher than 90%, which also shows that the optimal model saved basically meets the needs of intelligent analysis of the test results of the automatic tester. It can be seen from the accuracy of the category that the accuracy of fault types 1 (soft pressure plate configuration error), 2 (control word configuration error), and 3 (large current\voltage error) is lower than 90%. This embodiment uses a lightweight gradient boosting machine learning model (LGBM) to build a training test model and extract data such as the automatic test results of equipment and devices that apply automatic test technology. By applying these data to the model for training and analysis, an optimal model for judging the test data results is generated. The optimal model is saved and widely used in the analysis and judgment of automatic test results, significantly improving the intelligence level of substation automatic test technology.

[0075] In order to enable those skilled in the art to more clearly understand the technical solution of the present application, the implementation process of the automatic testing method for substation faults of the present application will be described in detail below in combination with specific embodiments.

[0076] This embodiment relates to a specific automatic testing method for substation faults, comprising the following steps:

[0077] Step S1: Obtain historical test parameters;

[0078] Specifically, the system obtains the historical test parameters of the substation through data acquisition equipment, including soft pressure plate parameters, control word parameters, zero-sequence overcurrent protection parameters and relay protection setting parameters.

[0079] Step S2: Obtain historical test results;

[0080] Specifically, the system obtains corresponding historical test results based on the data records of historical test parameters, including information such as fault type, fault time, and fault location.

[0081] Step S3: construct input vector and output vector;

[0082] Specifically, the system constructs an input vector based on the acquired historical test parameters and an output vector based on the historical test results. The input vector contains the setting values ​​of various parameters (0 represents the open state and 1 represents the closed state), and the output vector contains the fault code representing the fault.

[0083] Step S4: training a lightweight gradient boosting machine learning model;

[0084] Specifically, the system uses the constructed input vectors and output vectors to train the lightweight gradient boosting machine learning model. By determining multiple sets of initial model parameters, training the models separately and calculating the accuracy, the set of initial model parameters with the highest accuracy is finally selected as the optimal model parameters.

[0085] Step S5: Determine the type of fault that occurred in the substation;

[0086] Specifically, the system determines the type of fault currently occurring in the substation based on the model's prediction results and the mapping relationship between the fault type and the fault code. The system can also calculate the prediction accuracy of each fault type and adjust the model parameters based on the prediction accuracy to further optimize the accuracy of fault diagnosis.

[0087] Through the above steps, the system can automatically and accurately identify the fault type of the substation, reduce the occurrence of misjudgment, and improve the accuracy and efficiency of fault diagnosis.

[0088] The embodiment of the present application also provides an automatic test device for a transformer substation failure. It should be noted that the automatic test device for a transformer substation failure of the embodiment of the present application can be used to execute the automatic test method for transformer substation failure provided by the embodiment of the present application. The device is used to implement the above-mentioned embodiments and preferred implementation modes, and those that have been described will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware of a predetermined function. Although the device described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceived.

[0089] The following is an introduction to the automatic testing device for substation faults provided in an embodiment of the present application.

[0090] Figure 3 Schematic diagram of an automatic testing device for substation faults according to an embodiment of the present application. Figure 3 As shown, the device includes: an acquisition unit 10, an analysis unit 20 and a determination unit 30.

[0091] An acquisition unit 10 is used to acquire test parameters, wherein the test parameters are parameters for performing relay protection tests on the substation;

[0092] Specifically, the relay protection tester can simulate the intelligent terminal and the merging unit, and the general template file of the tester contains template information, protection test items and other parameter categories. According to the test results of the relay protection tester in the automatic test process and analysis of parameters that can be used, such as protection test phase setting, intelligent terminal switch status, merging unit current frequency, merging unit voltage frequency, merging unit voltage amplitude, etc., the system needs to obtain various parameters of the substation when conducting relay protection testing. These parameters may include electrical parameters such as voltage, current, temperature, frequency, as well as relay action time, protection setting, etc. These parameters are the basic data for fault diagnosis. By accurately obtaining these parameters, reliable input data can be provided for subsequent fault identification.

[0093] An analysis unit 20 is used to analyze the test parameters through a lightweight gradient boosting machine learning model to obtain a test result, wherein the lightweight gradient boosting machine learning model is obtained through machine learning training based on multiple sets of data, and each set of data in the multiple sets of data includes: historical test parameters and historical test results, and the test result is a result obtained by testing the substation;

[0094] Specifically, the model is trained based on a large amount of historical data, each set of which includes historical test parameters and corresponding historical test results. By learning from these data, the model can identify the complex relationship between test parameters and fault types, thereby generating prediction results for the current test.

[0095] The determination unit 30 is used to determine the type of fault that occurs in the substation according to the test result.

[0096] Specifically, the system determines the type of fault currently occurring in the substation based on the prediction results of the model. By analyzing the test results, the system can identify the specific fault type, such as soft pressure plate configuration error, control word configuration error, etc., and make corresponding responses or warnings.

[0097] This embodiment provides an automatic testing device for substation faults, which includes: an acquisition unit, an analysis unit, and a determination unit. The acquisition unit is used to acquire test parameters, wherein the test parameters are parameters for relay protection testing of the substation; the analysis unit is used to analyze the test parameters through a lightweight gradient boosting machine learning model to obtain test results, wherein the lightweight gradient boosting machine learning model is obtained through machine learning training based on multiple groups of data, and each group of data in the multiple groups of data includes: historical test parameters and historical test results, and the test results are the results obtained by testing the substation; the determination unit is used to determine the type of fault that occurred in the substation according to the test results. In this way, by intelligently analyzing the test parameters through the above-mentioned lightweight gradient boosting machine learning model, the type of fault that occurred in the substation can be automatically identified, the occurrence of misjudgment can be reduced, the accuracy and efficiency of fault diagnosis can be improved, and the problem of misjudgment in the automatic fault testing technology of substations in the prior art is solved.

[0098] In the specific implementation process, the acquisition unit includes: a history acquisition module and a model acquisition module.

[0099] A history acquisition module, used to acquire historical test parameters and historical test results, construct an input vector according to the historical test parameters, and construct an output vector according to the historical test results;

[0100] Specifically, the system first obtains a large number of historical test parameters and corresponding historical test results. Then, it constructs input vectors (historical test parameters) and output vectors (historical test results) based on these historical data. The input vector is used as input data for model training, and the output vector is used as a training target to help the model learn the relationship between parameters and fault types.

[0101] The model acquisition module is used to obtain an initial lightweight gradient boosting machine learning model, train the initial lightweight gradient boosting machine learning model through input vectors and output vectors to obtain optimal model parameters, configure the model parameters of the initial lightweight gradient boosting machine learning model to be the optimal model parameters, and obtain a lightweight gradient boosting machine learning model.

[0102] Specifically, the system obtains an initial lightweight gradient boosting machine learning model and trains the model through input vectors and output vectors. During the training process, the system adjusts the model parameters to achieve the best prediction effect. Finally, the model parameters are configured as the optimal model parameters, and a lightweight gradient boosting machine learning model that can be used for actual testing is obtained. It can be seen that the device can train a machine learning model that can accurately identify the type of substation faults, providing reliable support for subsequent fault diagnosis.

[0103] In order to obtain historical test parameters and historical test results, the history acquisition module includes: a parameter acquisition submodule and a result acquisition submodule.

[0104] The parameter acquisition submodule is used to obtain at least the soft pressure plate parameters, control word parameters, zero-sequence overcurrent protection parameters and relay protection setting parameters of the substation to obtain historical test parameters;

[0105] Specifically, the soft pressure plate is a control device in the substation, which is used to control the switching of the protection device. Obtaining the soft pressure plate parameters can understand the working status of the protection device; the control word parameters are key parameters in the substation control system, which are used to indicate various operations and states. Obtaining these parameters helps to understand the overall operating status of the system; zero-sequence overcurrent protection is a protection mechanism in the substation, which is used to detect ground faults. Obtaining these parameters can understand the performance of ground protection; the relay protection setting parameter is the action threshold of the relay under various fault conditions. Obtaining these parameters can understand the protection capability of the relay. By obtaining these key historical test parameters, the system provides the necessary data basis for subsequent model training. These parameters can reflect the operating status and protection performance of the substation, and are an important basis for fault diagnosis.

[0106] The result acquisition submodule is used to obtain the test results of historical test parameters and obtain historical test results.

[0107] Specifically, the device needs to obtain the test results corresponding to the above historical test parameters. These test results are usually fault records of substations in different states, including fault type, fault time, fault location and other information. Obtaining these test results can provide accurate output data for model training. It can be seen that by obtaining the test results of historical test parameters, the system can construct a complete training data set (input vector and output vector), providing a reliable basis for the training of lightweight gradient boosting machine learning models.

[0108] In order to construct the input vector and the output vector, the history acquisition module further includes: an input construction submodule and an output construction submodule.

[0109] An input construction submodule is used to construct an input vector according to different setting values ​​of historical test parameters, wherein the setting value includes 0 and 1, 0 represents an open state, and 1 represents a closed state;

[0110] Specifically, when a parameter is 0, it indicates that the corresponding device or protection mechanism is in the disconnected state; when a parameter is 1, it indicates that the corresponding device or protection mechanism is in the closed state or working state. By combining these setting values, the system can construct an input vector containing a variety of state information for model training. Different combinations of setting values ​​can reflect the operation status of the substation in various states and provide rich data support for fault diagnosis.

[0111] The output construction submodule is used to assign fault codes to the test results representing faults in the historical test results, and construct an output vector.

[0112] Specifically, the fault code can be a number or a character to identify different fault types, for example: 1 indicates a soft pressure plate configuration error, 2 a control word configuration error, 3 a fixed value configuration error, 4 indicates a large current\voltage error, etc. By combining these fault codes, the system can construct an output vector for the training target of the model. It can be seen that the device can convert the fault type in the test results into a standard fault code, which is convenient for model learning and identification. These fault codes can accurately reflect different fault types and provide clear targets for fault diagnosis.

[0113] Through the above steps, the system can obtain the historical test parameters and test results of the substation in detail and convert them into input vectors and output vectors suitable for model training. This process provides the necessary data support for the training of the lightweight gradient boosting machine learning model, ensuring that the model can accurately identify the fault type of the substation and reduce the occurrence of misjudgment.

[0114] In some embodiments, the model acquisition module includes: a determination submodule, a training submodule and an optimal determination submodule.

[0115] A determination submodule is used to determine multiple groups of initial model parameters, wherein each group of initial model parameters includes multiple model parameters, one of which is changed while the remaining model parameters remain unchanged;

[0116] Specifically, multiple groups of initial model parameters are first determined. Each group of initial model parameters includes multiple model parameters, one of which is changed, and the other model parameters remain unchanged. For example, if there are three model parameters A, B, and C, then one of A, B, and C can be changed respectively, while the other two parameters are kept unchanged, forming three groups of initial model parameters. The device can explore the combination of multiple model parameters, find the possible optimal parameter combination, and provide multiple starting points for subsequent model training.

[0117] The training submodule is used to train the initial lightweight gradient boosting machine learning model through input vectors and output vectors under each set of initial model parameters to obtain the accuracy of the training results;

[0118] Specifically, under each set of initial model parameters, the system uses the input vector (historical test parameters) and the output vector (historical test results) to train the initial lightweight gradient boosting machine learning model. After the training is completed, the system calculates the accuracy of the training results, that is, the model's prediction accuracy for the historical test data. Through this device, the system can evaluate the performance of the model under different combinations of model parameters, thereby screening out the parameter combination with the highest accuracy.

[0119] The optimal determination submodule is used to determine a set of initial model parameters with the highest accuracy as the optimal model parameters.

[0120] Specifically, among all the initial model parameter combinations, the system determines the set of initial model parameters with the highest accuracy as the optimal model parameters, and assigns them to the initial lightweight gradient boosting machine learning model, thereby obtaining a lightweight gradient boosting machine learning model that can be used for actual testing. Through this device, the system can ensure the optimization of model parameters and improve the prediction accuracy of the model in practical applications.

[0121] In order to determine the type of fault occurring in the substation, the determination unit includes: a relationship acquisition module and a fault determination module.

[0122] A relationship acquisition module is used to obtain a one-to-one mapping relationship between fault types and fault codes;

[0123] Specifically, first obtain a one-to-one mapping relationship between fault types and fault codes. For example, a short circuit fault corresponds to fault code 1, a ground fault corresponds to fault code 2, an overload fault corresponds to fault code 3, an undervoltage fault corresponds to fault code 4, etc. This step can establish a corresponding relationship between fault types and fault codes, providing a basis for subsequent fault type identification.

[0124] The fault determination module is used to determine the fault type corresponding to the fault code in the test result according to the one-to-one mapping relationship between the fault type and the fault code.

[0125] Specifically, the system parses the fault code in the model prediction result according to the mapping relationship between the acquired fault type and fault code to determine the corresponding fault type. For example, if the fault code in the model prediction result is 2, the system will determine that a ground fault has occurred in the substation. The device system can accurately identify the fault type of the substation and provide clear guidance for subsequent fault handling.

[0126] In some embodiments, the determination unit further includes: an adjustment module. The adjustment module is used to calculate the prediction accuracy under each fault type and adjust the optimal model parameters according to the prediction accuracy, wherein the prediction accuracy is the ratio of the number of correctly predicted data to the total number of predicted data.

[0127] This embodiment uses a lightweight gradient boosting machine learning model (LGBM) to build a training test model and extract data such as the automatic test results of equipment and devices that apply automatic test technology. By applying these data to the model for training and analysis, an optimal model for judging the test data results is generated. The optimal model is saved and widely used in the analysis and judgment of automatic test results, significantly improving the intelligence level of substation automatic test technology.

[0128] The automatic test device for substation faults includes a processor and a memory. The acquisition unit, analysis unit, and determination unit are stored in the memory as program units, and the processor executes the program units stored in the memory to implement corresponding functions. The modules are all located in the same processor; or, the modules are located in different processors in any combination.

[0129] The processor includes a kernel, and the kernel retrieves the corresponding program unit from the memory. One or more kernels can be provided, and the problem of misjudgment in the fault automatic testing technology of the substation in the prior art can be solved by adjusting the kernel parameters.

[0130] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0131] An embodiment of the present invention provides a computer-readable storage medium, the computer-readable storage medium including a stored program, wherein when the program is running, a device where the computer-readable storage medium is located is controlled to execute an automatic test device for a substation fault.

[0132] Specifically, the automatic test device for substation faults includes:

[0133] Step S201, obtaining test parameters, wherein the test parameters are parameters for performing relay protection tests on the substation;

[0134] Specifically, the relay protection tester can simulate the intelligent terminal and the merging unit, and the general template file of the tester contains template information, protection test items and other parameter categories. According to the test results of the relay protection tester in the automatic test process and analysis of parameters that can be used, such as protection test phase setting, intelligent terminal switch status, merging unit current frequency, merging unit voltage frequency, merging unit voltage amplitude, etc., the system needs to obtain various parameters of the substation when conducting relay protection testing. These parameters may include electrical parameters such as voltage, current, temperature, frequency, as well as relay action time, protection setting, etc. These parameters are the basic data for fault diagnosis. By accurately obtaining these parameters, reliable input data can be provided for subsequent fault identification.

[0135] Step S202, analyzing the test parameters through a lightweight gradient boosting machine learning model to obtain a test result, wherein the lightweight gradient boosting machine learning model is obtained through machine learning training based on multiple sets of data, each set of data in the multiple sets of data includes: historical test parameters and historical test results, and the test result is a result obtained by testing the substation;

[0136] Specifically, the model is trained based on a large amount of historical data, each set of which includes historical test parameters and corresponding historical test results. By learning from these data, the model can identify the complex relationship between test parameters and fault types, thereby generating prediction results for the current test.

[0137] Step S203: determine the type of fault that occurs in the substation according to the test result.

[0138] Specifically, the system determines the type of fault currently occurring in the substation based on the prediction results of the model. By analyzing the test results, the system can identify the specific fault type, such as soft pressure plate configuration error, control word configuration error, etc., and make corresponding responses or warnings.

[0139] Optionally, before obtaining the test parameters, the device also includes: obtaining historical test parameters and historical test results, constructing an input vector based on the historical test parameters, and constructing an output vector based on the historical test results; obtaining an initial lightweight gradient boosting machine learning model, training the initial lightweight gradient boosting machine learning model through the input vector and the output vector to obtain optimal model parameters, configuring the model parameters of the initial lightweight gradient boosting machine learning model to be the optimal model parameters, and obtaining a lightweight gradient boosting machine learning model.

[0140] Optionally, historical test parameters and historical test results are obtained, including: obtaining at least the soft pressure plate parameters, control word parameters, zero-sequence overcurrent protection parameters and relay protection setting parameters of the substation to obtain historical test parameters; obtaining test results of historical test parameters to obtain historical test results.

[0141] Optionally, an input vector is constructed according to historical test parameters, and an output vector is constructed according to historical test results, including: constructing an input vector according to different setting values ​​of historical test parameters, wherein the setting values ​​include 0 and 1, 0 represents an open state, and 1 represents a closed state; assigning fault codes to test results that characterize faults in historical test results to construct an output vector.

[0142] Optionally, an initial lightweight gradient boosting machine learning model is trained by input vectors and output vectors to obtain optimal model parameters, including: determining multiple sets of initial model parameters, wherein each set of initial model parameters includes multiple model parameters, wherein one model parameter changes and the remaining model parameters remain unchanged; under each set of initial model parameters, an initial lightweight gradient boosting machine learning model is trained by input vectors and output vectors to obtain the accuracy of the training results; and a set of initial model parameters with the highest accuracy is determined as the optimal model parameters.

[0143] Optionally, determining the fault type occurring in the substation according to the test result includes: obtaining a one-to-one mapping relationship between the fault type and the fault code; and determining the fault type corresponding to the fault code in the test result according to the one-to-one mapping relationship between the fault type and the fault code.

[0144] Optionally, after obtaining the test results, the device further includes: calculating the prediction accuracy under each fault type, and adjusting the optimal model parameters according to the prediction accuracy, wherein the prediction accuracy is the ratio of the number of correctly predicted data to the total number of predicted data.

[0145] An embodiment of the present invention provides an electronic device, including a processor, a memory, and a program stored in the memory and executable on the processor. When the processor executes the program, at least the following steps are implemented: step S201, obtaining test parameters, wherein the test parameters are parameters for performing relay protection tests on a substation; step S202, analyzing the test parameters through a lightweight gradient boosting machine learning model to obtain test results, wherein the lightweight gradient boosting machine learning model is obtained through machine learning training based on multiple groups of data, and each group of data in the multiple groups of data includes: historical test parameters and historical test results, and the test results are results obtained by testing the substation; step S203, determining the type of fault that occurred in the substation based on the test results.

[0146] Optionally, before obtaining the test parameters, the device also includes: obtaining historical test parameters and historical test results, constructing an input vector based on the historical test parameters, and constructing an output vector based on the historical test results; obtaining an initial lightweight gradient boosting machine learning model, training the initial lightweight gradient boosting machine learning model through the input vector and the output vector to obtain optimal model parameters, configuring the model parameters of the initial lightweight gradient boosting machine learning model to be the optimal model parameters, and obtaining a lightweight gradient boosting machine learning model.

[0147] Optionally, historical test parameters and historical test results are obtained, including: obtaining at least the soft pressure plate parameters, control word parameters, zero-sequence overcurrent protection parameters and relay protection setting parameters of the substation to obtain historical test parameters; obtaining test results of historical test parameters to obtain historical test results.

[0148] Optionally, an input vector is constructed according to historical test parameters, and an output vector is constructed according to historical test results, including: constructing an input vector according to different setting values ​​of historical test parameters, wherein the setting values ​​include 0 and 1, 0 represents an open state, and 1 represents a closed state; assigning fault codes to test results that characterize faults in historical test results to construct an output vector.

[0149] Optionally, an initial lightweight gradient boosting machine learning model is trained by input vectors and output vectors to obtain optimal model parameters, including: determining multiple sets of initial model parameters, wherein each set of initial model parameters includes multiple model parameters, wherein one model parameter changes and the remaining model parameters remain unchanged; under each set of initial model parameters, an initial lightweight gradient boosting machine learning model is trained by input vectors and output vectors to obtain the accuracy of the training results; and a set of initial model parameters with the highest accuracy is determined as the optimal model parameters.

[0150] Optionally, determining the fault type occurring in the substation according to the test result includes: obtaining a one-to-one mapping relationship between the fault type and the fault code; and determining the fault type corresponding to the fault code in the test result according to the one-to-one mapping relationship between the fault type and the fault code.

[0151] Optionally, after obtaining the test results, the device further includes: calculating the prediction accuracy under each fault type, and adjusting the optimal model parameters according to the prediction accuracy, wherein the prediction accuracy is the ratio of the number of correctly predicted data to the total number of predicted data.

[0152] The present application also provides a computer program product, including a computer program, which implements the steps of the device in each embodiment of the present application when executed by a processor: Step S201, obtaining test parameters, wherein the test parameters are parameters for performing relay protection tests on substations; Step S202, analyzing the test parameters through a lightweight gradient boosting machine learning model to obtain test results, wherein the lightweight gradient boosting machine learning model is obtained through machine learning training based on multiple sets of data, and each set of data in the multiple sets of data includes: historical test parameters and historical test results, and the test results are the results obtained by testing the substation; Step S203, determining the type of fault that occurred in the substation based on the test results. The device in this article can be a server, PC, PAD, mobile phone, etc.

[0153] Optionally, before obtaining the test parameters, the device also includes: obtaining historical test parameters and historical test results, constructing an input vector based on the historical test parameters, and constructing an output vector based on the historical test results; obtaining an initial lightweight gradient boosting machine learning model, training the initial lightweight gradient boosting machine learning model through the input vector and the output vector to obtain optimal model parameters, configuring the model parameters of the initial lightweight gradient boosting machine learning model to be the optimal model parameters, and obtaining a lightweight gradient boosting machine learning model.

[0154] Optionally, historical test parameters and historical test results are obtained, including: obtaining at least the soft pressure plate parameters, control word parameters, zero-sequence overcurrent protection parameters and relay protection setting parameters of the substation to obtain historical test parameters; obtaining test results of historical test parameters to obtain historical test results.

[0155] Optionally, an input vector is constructed according to historical test parameters, and an output vector is constructed according to historical test results, including: constructing an input vector according to different setting values ​​of historical test parameters, wherein the setting values ​​include 0 and 1, 0 represents an open state, and 1 represents a closed state; assigning fault codes to test results that characterize faults in historical test results to construct an output vector.

[0156] Optionally, an initial lightweight gradient boosting machine learning model is trained by input vectors and output vectors to obtain optimal model parameters, including: determining multiple sets of initial model parameters, wherein each set of initial model parameters includes multiple model parameters, wherein one model parameter changes and the remaining model parameters remain unchanged; under each set of initial model parameters, an initial lightweight gradient boosting machine learning model is trained by input vectors and output vectors to obtain the accuracy of the training results; and a set of initial model parameters with the highest accuracy is determined as the optimal model parameters.

[0157] Optionally, determining the fault type occurring in the substation according to the test result includes: obtaining a one-to-one mapping relationship between the fault type and the fault code; and determining the fault type corresponding to the fault code in the test result according to the one-to-one mapping relationship between the fault type and the fault code.

[0158] Optionally, after obtaining the test results, the device further includes: calculating the prediction accuracy under each fault type, and adjusting the optimal model parameters according to the prediction accuracy, wherein the prediction accuracy is the ratio of the number of correctly predicted data to the total number of predicted data.

[0159] Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order than here, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0160] Those skilled in the art will appreciate that the embodiments of the present application may be provided as devices, systems, or computer program products. Therefore, the present application may adopt the form of complete hardware embodiments, complete software embodiments, or embodiments in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0161] The present application is described with reference to the flowcharts and / or block diagrams of the apparatus, device (system), and computer program product according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0162] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0163] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0164] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0165] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0166] Computer readable media include permanent and non-permanent, removable and non-removable media that can be used to store information by any device or technology. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0167] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, device, commodity or equipment including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, device, commodity or equipment. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, device, commodity or equipment including the elements.

[0168] From the above description, it can be seen that the above embodiments of the present application achieve the following technical effects:

[0169] 1) The present application provides an automatic testing device for substation faults. First, the test parameters are obtained, wherein the test parameters are parameters for relay protection testing of the substation; the test parameters are analyzed by a lightweight gradient boosting machine learning model to obtain test results, wherein the lightweight gradient boosting machine learning model is obtained by machine learning training based on multiple groups of data, and each group of data in the multiple groups of data includes: historical test parameters and historical test results, and the test results are the results obtained by testing the substation; the type of fault that occurred in the substation is determined according to the test results. In this way, by intelligently analyzing the test parameters through the above-mentioned lightweight gradient boosting machine learning model, the type of fault that occurred in the substation can be automatically identified, the occurrence of misjudgment can be reduced, the accuracy and efficiency of fault diagnosis can be improved, and the problem of misjudgment in the automatic fault testing technology of substations in the prior art is solved.

[0170] 2) The present application provides an automatic testing device for substation faults, which includes: an acquisition unit, an analysis unit, and a determination unit. The acquisition unit is used to acquire test parameters, wherein the test parameters are parameters for relay protection testing of the substation; the analysis unit is used to analyze the test parameters through a lightweight gradient boosting machine learning model to obtain test results, wherein the lightweight gradient boosting machine learning model is obtained through machine learning training based on multiple groups of data, and each group of data in the multiple groups of data includes: historical test parameters and historical test results, and the test results are the results obtained by testing the substation; the determination unit is used to determine the type of fault that occurred in the substation based on the test results. In this way, by intelligently analyzing the test parameters through the above-mentioned lightweight gradient boosting machine learning model, the type of fault that occurred in the substation can be automatically identified, the occurrence of misjudgment can be reduced, the accuracy and efficiency of fault diagnosis can be improved, and the problem of misjudgment in the automatic fault testing technology of substations in the prior art is solved.

[0171] The above are only preferred embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. An automatic testing method for substation faults, characterized in that: include: Acquiring test parameters, wherein the test parameters are parameters for performing relay protection tests on the substation; The test parameters are analyzed by a lightweight gradient boosting machine learning model to obtain a test result, wherein the lightweight gradient boosting machine learning model is obtained by machine learning training based on multiple sets of data, each set of data in the multiple sets of data includes: historical test parameters and historical test results, and the test results are results obtained by testing the substation; The type of fault occurring in the substation is determined according to the test result.

2. The automatic testing method for substation fault according to claim 1 is characterized in that: Before obtaining the test parameters, the method further includes: Acquire the historical test parameters and the historical test results, construct an input vector according to the historical test parameters, and construct an output vector according to the historical test results; Obtain an initial lightweight gradient boosting machine learning model, train the initial lightweight gradient boosting machine learning model through the input vector and the output vector to obtain optimal model parameters, configure the model parameters of the initial lightweight gradient boosting machine learning model to the optimal model parameters, and obtain the lightweight gradient boosting machine learning model.

3. The automatic testing method for substation fault according to claim 2 is characterized in that: Acquiring the historical test parameters and the historical test results includes: At least obtaining the soft pressure plate parameters, control word parameters, zero-sequence overcurrent protection parameters and relay protection setting parameters of the substation to obtain the historical test parameters; Acquire the test results of the historical test parameters to obtain the historical test results.

4. The automatic testing method for substation fault according to claim 2 is characterized in that: Constructing an input vector according to the historical test parameters, and constructing an output vector according to the historical test results, including: Constructing the input vector according to different setting values ​​of the historical test parameter, wherein the setting value includes 0 and 1, 0 represents an open state, and 1 represents a closed state; A fault code is assigned to the test result representing the fault in the historical test results to construct the output vector.

5. The automatic testing method for substation fault according to claim 2, characterized in that: Training the initial lightweight gradient boosting machine learning model by using the input vector and the output vector to obtain optimal model parameters includes: Determine multiple groups of initial model parameters, wherein each group of the initial model parameters includes multiple model parameters, one of the model parameters is changed, and the remaining model parameters remain unchanged; Under each set of the initial model parameters, training the initial lightweight gradient boosting machine learning model through the input vector and the output vector to obtain the accuracy of the training result; A set of the initial model parameters with the highest accuracy is determined as the optimal model parameters.

6. The automatic testing method for substation fault according to claim 1, characterized in that: Determining the type of fault occurring in the substation according to the test result includes: Get the one-to-one mapping relationship between fault type and fault code; The fault type corresponding to the fault code in the test result is determined according to the one-to-one mapping relationship between the fault type and the fault code.

7. The automatic testing method for substation fault according to claim 1, characterized in that: After obtaining the test results, the method further includes: The prediction accuracy under each fault type is calculated, and the optimal model parameters are adjusted according to the prediction accuracy, wherein the prediction accuracy is the ratio of the number of correctly predicted data to the total number of predicted data.

8. An automatic testing device for substation faults, characterized in that: include: An acquisition unit, used for acquiring test parameters, wherein the test parameters are parameters for performing relay protection tests on the substation; An analysis unit is used to analyze the test parameters through a lightweight gradient boosting machine learning model to obtain a test result, wherein the lightweight gradient boosting machine learning model is obtained through machine learning training based on multiple sets of data, each set of data in the multiple sets of data includes: historical test parameters and historical test results, and the test results are results obtained by testing the substation; A determination unit is used to determine the type of fault that occurred in the substation according to the test result.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the automatic testing method for substation faults according to any one of claims 1 to 7.

10. An electronic device, characterized in that: include: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include a method for executing the automatic testing method of the substation fault according to any one of claims 1 to 7.

Images