Transformer substation fault diagnosis method and device, electronic equipment and storage medium

Abstract

The invention provides a transformer substation fault diagnosis method and device, electronic equipment and a storage medium, and relates to the technical field of computers, and the method comprises the steps: obtaining fault signal data of all equipment in a target transformer substation; inputting the fault signal data of each device into a fault diagnosis model to obtain a fault association relationship among the fault signal data output by the fault diagnosis model and a fault diagnosis result of the target substation; respectively performing expert checking on the fault association relationship and the fault diagnosis result to generate an association relationship checking result and a fault diagnosis checking result; storing the fault signal data, the correlation checking result and the fault diagnosis checking result to a fault diagnosis knowledge base; the fault diagnosis model is obtained based on training of the fault diagnosis knowledge base. According to the method and the device provided by the invention, the efficiency and the accuracy of transformer substation fault diagnosis are improved.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a substation fault diagnosis method, device, electronic equipment and storage medium. Background Art

[0002] With the continuous development of the power system, substations, as important power equipment nodes, have seen an increasing frequency of equipment failures. Especially during equipment operation, the massive signal data generated increases the difficulty of fault diagnosis.

[0003] Traditional fault diagnosis methods rely on manual experience for judgment and analysis, which are often inefficient and easily affected by human factors.

[0004] Therefore, how to improve the efficiency and accuracy of substation fault diagnosis has become a technical problem that needs to be solved urgently in the industry. Summary of the invention

[0005] The present invention provides a substation fault diagnosis method, device, electronic equipment and storage medium, which are used to solve the technical problem of how to improve the efficiency and accuracy of substation fault diagnosis.

[0006] The present invention provides a substation fault diagnosis method, comprising: Obtain fault signal data of each device in the target substation; Inputting the fault signal data of each device into a fault diagnosis model to obtain the fault association relationship between the fault signal data output by the fault diagnosis model and the fault diagnosis result of the target substation; Performing expert verification on the fault association relationship and the fault diagnosis result respectively to generate an association verification result and a fault diagnosis verification result; Storing the fault signal data, the association relationship verification result and the fault diagnosis verification result in a fault diagnosis knowledge base; Wherein, the fault diagnosis model is obtained by training based on the fault diagnosis knowledge base.

[0007] In some embodiments, the fault diagnosis model includes a clustering analysis layer and a result recognition layer connected in sequence.

[0008] In some embodiments, the inputting of the fault signal data of each device into the fault diagnosis model to obtain the fault association relationship between the fault signal data output by the fault diagnosis model and the fault diagnosis result of the target substation includes: Inputting the fault signal data of each device into the cluster analysis layer, and clustering the fault signal data of each device by the cluster analysis layer to obtain a plurality of clusters; the cluster includes at least one fault signal data; Input the multiple clusters into the result recognition layer, and the result recognition layer identifies the fault signal data in each cluster, obtains the fault association relationships between the fault signal data in each cluster and the fault diagnosis results of each cluster, and determines the fault diagnosis result of the target substation based on the fault diagnosis results of each cluster.

[0009] In some embodiments, the inputting the fault signal data of each device into the clustering analysis layer, and the clustering analysis layer clustering the fault signal data of each device to obtain multiple clusters includes: Determine the initial clustering centers in the current iteration process; Determine the distances between each fault signal data and the clustering centers of each cluster; Assign each fault signal data to the clustering center with the minimum distance to obtain the updated clusters; Determine the updated clustering centers based on the means of the fault signal data in the updated clusters; In the case that the distance between the updated clustering centers and the initial clustering centers is greater than a preset distance or the number of iterations is less than a preset number of times, use the updated clustering centers as the initial clustering centers in the next iteration process and continue the iteration; In the case that the distance between the updated clustering centers and the initial clustering centers is less than a preset distance or the number of iterations is greater than a preset number of times, stop the iteration.

[0010] In some embodiments, the determining the distances between each fault signal data and the clustering centers of each cluster includes: Determine the calculation weights of each data item in each fault signal data; Based on the calculation weights of each data item and the differences between each data item in each fault signal data and each data item in the clustering center, determine the distances between each fault signal data and the clustering centers of each cluster.

[0011] In some embodiments, the method further includes: Determine that the current fault signal data is different from the historical fault signal data; Store the current fault signal data, as well as the corresponding association relationship verification result and fault diagnosis verification result of the current fault signal data, into the fault diagnosis knowledge base; Based on the current fault signal data, as well as the corresponding association relationship verification result and fault diagnosis verification result of the current fault signal data, perform incremental training on the fault diagnosis model.

[0012] In some embodiments, the obtaining the fault signal data of each device in the target substation includes: Obtain the initial signals of each device sent by the data acquisition and monitoring control system, fault recording system, and centralized control system of the target substation; Extract features from the initial signals to obtain the fault signal data of each device; the fault signal data includes device type, fault occurrence time, fault duration, device operating status, and protection device actions; the device operating status includes current, voltage, and temperature.

[0013] In some embodiments, after extracting features from the initial signals to obtain the fault signal data of each device, the method further includes: Preprocess the fault signal data of each device; the preprocessing includes at least one of data cleaning, data standardization, and data normalization.

[0014] The present invention provides a substation fault diagnosis device, including: An acquisition module, configured to acquire the fault signal data of each device in the target substation; A diagnosis module, configured to input the fault signal data of each device into a fault diagnosis model to obtain the fault correlation relationship between the fault signal data output by the fault diagnosis model and the fault diagnosis result of the target substation; A verification module, configured to respectively perform expert verification on the fault correlation relationship and the fault diagnosis result to generate a correlation relationship verification result and a fault diagnosis verification result; A storage module, configured to store the fault signal data, the correlation relationship verification result, and the fault diagnosis verification result into a fault diagnosis knowledge base; Wherein, the fault diagnosis model is trained based on the fault diagnosis knowledge base.

[0015] The present invention provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the substation fault diagnosis method is implemented.

[0016] The present invention provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the substation fault diagnosis method is implemented.

[0017] The present invention provides a computer program product, including a computer program. When the computer program is executed by a processor, the substation fault diagnosis method is implemented.

[0018] The substation fault diagnosis method, device, electronic device and storage medium provided by the embodiments of the present invention obtain the fault signal data of each device in the target substation; input the fault signal data of each device into the fault diagnosis model to obtain the fault correlation relationship between the fault signal data output by the fault diagnosis model and the fault diagnosis result of the target substation; respectively conduct expert verification on the fault correlation relationship and the fault diagnosis result to generate the correlation relationship verification result and the fault diagnosis verification result; store the fault signal data, the correlation relationship verification result and the fault diagnosis verification result in the fault diagnosis knowledge base; since the fault diagnosis knowledge base is constructed by collecting the fault signal data, the correlation relationship verification result and the fault diagnosis verification result, the fault diagnosis model is trained according to the fault diagnosis knowledge base, and the fault diagnosis model deeply analyzes the fault signal data to determine the fault correlation relationship and the fault diagnosis result, which can automatically process a large amount of fault signal data without relying on manual experience for judgment and analysis, improving the efficiency and accuracy of substation fault diagnosis and providing intelligent support for the operation and maintenance work of the substation. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The accompanying drawings herein are incorporated into the specification and form a part of the specification, showing embodiments consistent with the present invention and used together with the specification to explain the principles of the present invention.

[0020] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the accompanying drawings required for use in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are some embodiments of the present invention, and for those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0021] Figure 1 is one of the flow diagrams of the substation fault diagnosis method provided by the present invention.

[0022] Figure 2 is the structural diagram of the fault diagnosis model provided by the present invention.

[0023] Figure 3 is the second flow diagram of the substation fault diagnosis method provided by the present invention.

[0024] Figure 4 is the structural diagram of the substation fault diagnosis device provided by the present invention.

[0025] Figure 5 is the structural diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0026] To enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0027] It should be noted that the terms "first", "second", etc. in the present invention are used to distinguish similar objects, and do not necessarily need to describe a specific order or sequence. It should be understood that such used data can be interchanged under appropriate circumstances so that the embodiments of the present invention described here can be implemented in an order other than those illustrated or described here. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units or modules does not necessarily have to be limited to those steps or units or modules clearly listed, but may include other steps or units or modules not clearly listed or inherent to these processes, methods, products or devices.

[0028] ‌Due to the complexity, non-linearity and multi-dimensionality of the substation equipment fault signal data, how to effectively extract and mine the correlation relationship between fault signals and apply it to actual fault prediction and diagnosis is still a challenging problem.

[0029] To solve the above technical problems, Figure 1 is one of the schematic flowcharts of the substation fault diagnosis method provided by the present invention, as Figure 1 shown, the method includes step 110, step 120, step 130 and step 140.

[0030] Step 110: Obtain the fault signal data of each device in the target substation.

[0031] Specifically, the execution subject of the substation fault diagnosis method provided by the embodiments of the present invention is a substation fault diagnosis device. This device can be implemented by software, such as a substation fault diagnosis program running on a computer; or can be implemented by hardware, such as a computer or a server that executes the substation fault diagnosis method, etc.

[0032] A substation is a power facility in the power system that transforms voltage, receives and distributes electric energy, controls the flow of electric power, and adjusts voltage. It connects power grids at different voltage levels through transformers and is an important part of the power system. The equipment in a substation includes transformers, transmission lines, busbars, circuit breakers, disconnectors, and protection devices, etc. In actual operation, faults such as transformer faults, busbar faults, circuit breaker faults, short-circuit faults, and grounding faults are likely to occur in substations.

[0033] The target substation refers to the substation that needs to conduct fault diagnosis. Fault signal data refers to the data that can reflect the status information and characteristic parameters of various equipment in the substation when a fault or abnormal situation occurs.

[0034] Communication can be established between the substation fault diagnosis device and the Supervisory Control And Data Acquisition (SCADA) system, fault recording system, and centralized control system of the target substation to obtain the initial signals of each device. The initial signal is the original signal collected from each system that reflects the equipment fault.

[0035] Feature extraction is performed on the initial signal to obtain the fault signal data of each device. The data items in the fault signal data at least include equipment type, fault occurrence time, fault duration, equipment operation status, and protection device operation. The equipment operation status can include current, voltage, and temperature. For example, the fault signal data of each device can be represented by a matrix of size where represents the number of fault signal data, and represents the number of data items (feature dimensions), as shown in Table 1.

[0036] Table 1 Fault Signal Data

[0037] After obtaining the fault signal data of each device, the fault signal data can be preprocessed. Data preprocessing can include data cleaning, data standardization, and data normalization, etc.

[0038] Data cleaning can include handling missing values, outliers, and duplicate values. Statistical measures such as mean, median, and mode can be used to fill in missing values, or based on the correlation between data, predictive models such as regression and decision trees can be adopted to fill in missing values. For detected outliers (e.g., data greater than 1.5 times the rated value), deletion, correction, or retention can be selected according to the specific situation. If the outlier is caused by an error, it can be corrected; if the reason cannot be determined but it has a significant impact on the analysis results, deletion can be considered; if the outlier itself has special significance or may contain valuable information, it can be retained and specially processed in subsequent analysis. Duplicate values increase the data volume and affect the analysis efficiency and result accuracy. It can be processed by data deduplication methods, such as deleting duplicate records according to the unique identifier of the data or specific key fields.

[0039] Data standardization refers to transforming data so that it has zero mean and unit variance. The Z-score standardization method can be adopted. By calculating the mean and standard deviation of the data, the data is converted into the form of a standard normal distribution, which can be expressed by the formula: .

[0040] Where, is the data before standardization, is the data after standardization, is the feature 's mean, is the feature 's standard deviation.

[0041] Data normalization refers to scaling data to a specific interval, usually [0,1]. By finding the maximum and minimum values of the data, the data is linearly transformed into the specified interval.

[0042] Step 120: Input the fault signal data of each device into the fault diagnosis model to obtain the fault correlation relationship between the fault signal data output by the fault diagnosis model and the fault diagnosis result of the target substation.

[0043] Specifically, the fault correlation relationship refers to the internal connection and mutual influence existing between different fault signal data. These fault correlation relationships help to more accurately diagnose faults, locate the fault source, and predict the fault development trend. The fault correlation relationship can include temporal relationship, causal relationship, and device correlation relationship.

[0044] The temporal relationship means that some fault signals will appear in a certain time sequence. For example, when a short-circuit fault occurs inside a transformer, first a signal of sudden increase in current will appear, followed by a signal of sharp rise in temperature. If the fault further intensifies, it will also cause the protection device to act and send a tripping signal, etc.

[0045] Causality means that the failure of one device may be the direct cause of the failure of another device. For example, when a short - circuit fault occurs in a busbar, it will cause over - current, over - voltage and other fault signals to appear in multiple devices connected to the busbar. At this time, the short - circuit fault of the busbar is the cause, and the over - current and over - voltage signals of other devices are the effects.

[0046] The device association relationship means that each device forms a whole through electrical connection. When a certain device fails, it will affect the operating state of other devices electrically connected to it, and then generate corresponding fault signals.

[0047] The fault diagnosis result refers to the possible fault types that may occur in the target substation. The fault diagnosis result is the root cause of the faults that occur in multiple devices in the target substation.

[0048] The fault signal data of each device can be subjected to feature extraction and feature learning through machine learning or deep learning methods to determine the fault association relationship and the fault diagnosis result.

[0049] A fault diagnosis model can be constructed with a neural network as the initial model. The neural network can be selected from convolutional neural networks, recurrent neural networks, generative adversarial networks, attention mechanism networks, etc.

[0050] A large amount of historical fault signal data of each device can be collected as samples, and the fault association relationship and fault diagnosis result corresponding to these historical fault signal data are used as labels to train the initial model, improve the feature learning ability of the initial model and the prediction ability of the fault association relationship and fault diagnosis result, and finally obtain the fault diagnosis model.

[0051] After inputting the fault signal data of each device into the trained fault diagnosis model, the fault diagnosis model can output the fault association relationship between the fault signal data and the fault diagnosis result of the target substation.

[0052] Step 130: Expertly check the fault association relationship and the fault diagnosis result respectively to generate the association relationship check result and the fault diagnosis check result.

[0053] Specifically, an expert system refers to an artificial intelligence system that uses the knowledge and experience mastered by human experts, simulates the expert decision - making ability through a computer program, and solves complex problems in a specific field.

[0054] The fault association relationship and the fault diagnosis result output by the fault diagnosis model can be sent to the operator or the expert system. The operator or the expert system, based on experience, analyzes the relationship between the signal and the fault, checks the fault association relationship and the fault diagnosis result respectively, and generates the association relationship check result and the fault diagnosis check result.

[0055] If the fault association relationship output by the fault diagnosis model is consistent with the analysis result of the expert system, the fault association relationship output by the model can be used as the result of the association relationship verification; if not, the analysis result of the expert system is determined as the result of the association relationship verification. The processing process of the fault diagnosis verification result is similar to this, and will not be elaborated here.

[0056] The expert system verifies and optimizes the fault association relationship and the fault diagnosis result, ensuring the reliability and accuracy of the association rules and avoiding misjudgment of the algorithm result.

[0057] Step 140: Store the fault signal data, the result of the association relationship verification, and the result of the fault diagnosis verification into the fault diagnosis knowledge base; among them, the fault diagnosis model is trained based on the fault diagnosis knowledge base.

[0058] Specifically, since there is interpretability among the fault signal data, the result of the association relationship verification, and the result of the fault diagnosis verification, a fault diagnosis knowledge base can be constructed. The fault signal data, the result of the association relationship verification, and the result of the fault diagnosis verification are used as an association rule (also a knowledge item) for fault diagnosis and stored in the fault diagnosis knowledge base.

[0059] The fault diagnosis knowledge base centrally stores a large amount of fault signal data and diagnostic knowledge. The diagnosis system or personnel can quickly query and match, reducing the diagnosis time and improving the efficiency. Especially when dealing with complex faults, the advantage is more obvious. Based on a large amount of actual data and the expert verification result, the knowledge base can provide an accurate basis for diagnosis, enabling the system or personnel to more accurately identify faults, reducing misjudgment and missed judgment, and ensuring the reliability of diagnosis. The stored data provides a basis for the update and optimization of the knowledge base. Regularly analyzing and evaluating the data can find inaccurate or outdated knowledge and update and optimize it in a timely manner to ensure that the knowledge base adapts to the changes of the equipment and the application of new technologies and maintains high efficiency.

[0060] The association rules in the fault diagnosis knowledge base can also be used as samples to train the fault diagnosis model, improving the accuracy of the fault diagnosis model and achieving more accurate diagnosis and prediction.

[0061] The substation fault diagnosis method provided by the embodiments of the present invention obtains the fault signal data of each device in the target substation; inputs the fault signal data of each device into the fault diagnosis model to obtain the fault correlation relationship between the fault signal data output by the fault diagnosis model and the fault diagnosis result of the target substation; respectively conducts expert verification on the fault correlation relationship and the fault diagnosis result to generate the correlation relationship verification result and the fault diagnosis verification result; stores the fault signal data, the correlation relationship verification result and the fault diagnosis verification result in the fault diagnosis knowledge base; since the fault diagnosis knowledge base is constructed by collecting the fault signal data, the correlation relationship verification result and the fault diagnosis verification result, the fault diagnosis model is trained according to the fault diagnosis knowledge base, and the fault diagnosis model deeply analyzes the fault signal data to determine the fault correlation relationship and the fault diagnosis result, which can automatically process a large amount of fault signal data, without relying on manual experience for judgment and analysis, improves the efficiency and accuracy of substation fault diagnosis, and provides intelligent support for the operation and maintenance work of the substation.

[0062] It should be noted that each embodiment of the present invention can be freely combined, the order can be swapped or executed independently, and does not need to rely on a fixed execution order.

[0063] In some embodiments, the fault diagnosis model includes a clustering analysis layer and a result recognition layer connected in sequence.

[0064] Specifically, Figure 2 is the structural schematic diagram of the fault diagnosis model provided by the present invention. As Figure 2 shown, the fault diagnosis model 200 includes a clustering analysis layer 210 and a result recognition layer 220. The clustering analysis layer is used to perform clustering analysis on the fault signal data of each device, and the result recognition layer is used to recognize the clustering analysis result to determine the fault correlation relationship and the fault diagnosis result.

[0065] In some embodiments, inputting the fault signal data of each device into the fault diagnosis model to obtain the fault correlation relationship between the fault signal data output by the fault diagnosis model and the fault diagnosis result of the target substation includes: Input the fault signal data of each device into the clustering analysis layer, and the clustering analysis layer clusters the fault signal data of each device to obtain multiple clusters; a cluster includes at least one fault signal data; Input the multiple clusters into the result recognition layer, and the result recognition layer recognizes the fault signal data in each cluster to obtain the fault correlation relationship between the fault signal data in each cluster and the fault diagnosis result of each cluster, and determines the fault diagnosis result of the target substation based on the fault diagnosis result of each cluster.

[0066] Specifically, the clustering algorithm is an important method of unsupervised learning. It can divide the data points in a dataset into different clusters or groups, such that the data points within the same cluster have a relatively high similarity, while the data points between different clusters have a relatively low similarity. Clustering algorithms include the K-means algorithm, hierarchical clustering algorithm, density clustering algorithm, fuzzy clustering algorithm, etc.

[0067] The clustering analysis layer can perform clustering analysis on the fault signal data of each device, and allocate the fault signal data with similar characteristics to the same cluster. The result of the clustering analysis is to obtain multiple clusters, and each cluster includes at least one fault signal data. The fault signal data within the same cluster have a high similarity and may indicate the same fault type.

[0068] The result recognition layer identifies each cluster output by the clustering analysis layer, identifies the fault association relationships between the fault signal data in each cluster, and identifies the fault diagnosis results of each cluster. After obtaining the fault diagnosis results of each cluster, the result recognition layer also comprehensively analyzes the fault diagnosis results of each cluster to determine the fault diagnosis result of the target substation. For example, the fault diagnosis results of each cluster can be statistically analyzed, and the fault diagnosis result with the largest number of clusters can be determined as the fault diagnosis result of the target substation.

[0069] The substation fault diagnosis method provided by the embodiments of the present invention first performs clustering analysis on the fault signal data through a fault diagnosis model, and then identifies the clustering analysis results, simplifies the data scale, saves computing resources for subsequent identification, and improves the processing speed; clustering can separate noise data from normal data, and the noise impact can be eliminated or reduced during subsequent identification, improving the accuracy of substation fault diagnosis.

[0070] In some embodiments, inputting the fault signal data of each device into the clustering analysis layer, and the clustering analysis layer performs clustering on the fault signal data of each device to obtain multiple clusters, including: Determine the initial clustering centers in the current iteration process; Determine the distances between each fault signal data and the clustering centers of each cluster; Allocate each fault signal data to the clustering center with the minimum distance to obtain the updated clusters; Based on the mean values of the fault signal data in the updated clusters, determine the updated clustering centers; In the case that the distance between the updated clustering centers and the initial clustering centers is greater than a preset distance or the number of iterations is less than a preset number of times, use the updated clustering centers as the initial clustering centers in the next iteration process and continue the iteration; In the case that the distance between the updated clustering centers and the initial clustering centers is less than a preset distance or the number of iterations is greater than a preset number of times, stop the iteration.

[0071] Specifically, the clustering analysis layer can use the K-means algorithm for clustering analysis, including multiple iteration processes. Before the start of the current iteration process, the initial clustering centers need to be determined.

[0072] If the current iteration process is the first iteration process, several fault signal data can be randomly selected as the initial clustering centers. For example, it can be considered that the fault types of a substation are divided into 3 categories, namely circuit breaker short-circuit fault, capacitor overvoltage fault, and protection device malfunction fault. Then, 1 fault signal data corresponding to these 3 types of faults can be randomly selected as the initial clustering centers. If the current iteration process is not the first iteration process, the clustering centers generated in the previous iteration process can be determined as the initial clustering centers in the current iteration process.

[0073] The distances between each fault signal data and the clustering centers of each cluster can be calculated according to algorithms such as Euclidean distance, and then each fault signal data can be assigned to the clustering center with the smallest distance, so as to obtain the updated clusters.

[0074] After all the fault signal data are assigned to their corresponding clustering centers, the clustering centers of each cluster are recalculated, that is, the mean value of all the fault signal data in each cluster is calculated as the updated clustering center. In this way, the updated clustering centers gradually approach the true center of the data.

[0075] The distance between the updated clustering center and the initial clustering center can be calculated according to Euclidean distance, etc.: If this distance is greater than the preset distance or the number of iterations in the current iteration process is less than the preset number of times, it means that the updated clustering center is not yet the true center of the data, and iterative optimization can still be performed. The updated clustering center can be used as the initial clustering center in the next iteration process and continue to iterate.

[0076] If this distance is less than the preset distance or the number of iterations is greater than the preset number of times, it means that the updated clustering center is already the true center of the data, or the clustering center is not the true center of the data but the clustering center no longer changes, and the iteration can be stopped.

[0077] Finally, 3 clusters are obtained, and each cluster indicates a fault type. The equipment fault signals in each cluster have similar characteristics. For example, the fault signals including buses, reactors, and capacitors show similarity in temperature anomaly values and current anomaly values, which may imply that their faults are related to harmonics or overheating in the power grid.

[0078] The substation fault diagnosis method provided by the embodiments of the present invention clusters the fault signal data of each device to obtain multiple clusters, which can separate the noise data from the normal data. During subsequent identification, the influence of noise can be eliminated or reduced, improving the accuracy of substation fault diagnosis and reducing the workload of manual analysis.

[0079] In some embodiments, determining the distance between each fault signal data and the cluster center of each cluster includes: Determining the calculation weight of each data item in each fault signal data; Based on the calculation weight of each data item and the difference between each data item in each fault signal data and each data item in the cluster center, determining the distance between each fault signal data and the cluster center of each cluster.

[0080] Specifically, in the fault signal data, the influence degree of each data item on the fault diagnosis result is different. Therefore, the calculation weight of each data item can be determined according to the influence degree of each data item. The greater the influence degree, the greater the calculation weight.

[0081] Determining the distance between each fault signal data and the cluster center of each cluster according to the calculation weight of each data item and the difference between each data item in each fault signal data and each data item in the cluster center can be expressed by the formula: In the formula, represents the th fault signal data and the th cluster center the distance between; represents the calculation weight of the first data item, represents the th calculation weight of the data item, represents the fault signal data the difference between the first data item in and the first data item in the cluster center ; represents the fault signal data the th data item in and the cluster center the th data item in the difference. The fault signal data contains a total of data items.

[0082] When the substation fault diagnosis method provided by the embodiments of the present invention determines the distance between each fault signal data and the cluster center of each cluster, it considers the influence degree of each data item on the fault diagnosis result, improving the accuracy of substation fault diagnosis.

[0083] In some embodiments, the method further includes: Determining that the current fault signal data is different from the historical fault signal data; Storing the current fault signal data, as well as the associated relationship verification result and the fault diagnosis verification result corresponding to the current fault signal data, into the fault diagnosis knowledge base; Based on the current fault signal data, as well as the associated relationship verification result and the fault diagnosis verification result corresponding to the current fault signal data, performing incremental training on the fault diagnosis model.

[0084] Specifically, incremental training means that, based on the existing training model, new training data is continuously added, and the model is updated and optimized, rather than retraining the entire model from scratch.

[0085] After performing clustering analysis and identification on the current fault signal data, the current fault signal data can be compared with the historical fault signal data. If the current fault signal data is different from the historical fault signal data, it indicates that the current fault signal data is new fault signal data.

[0086] On the one hand, the current fault signal data, as well as the associated relationship verification result and the fault diagnosis verification result corresponding to the current fault signal data, can be stored in the fault diagnosis knowledge base. On the other hand, an automatic update mechanism can be set up to perform incremental training on the fault diagnosis model at a set time or according to specific trigger conditions, using the current fault signal data, as well as the associated relationship verification result and the fault diagnosis verification result corresponding to the current fault signal data, so as to optimize and adjust the parameters of the fault diagnosis model.

[0087] The substation fault diagnosis method provided by the embodiments of the present invention can continuously adapt to new fault signal data by updating the knowledge base and the fault diagnosis model, continuously improve the accuracy and efficiency of fault diagnosis, and improve the accuracy of substation fault diagnosis.

[0088] Figure 3 is the second flow schematic diagram of the substation fault diagnosis method provided by the present invention. As Figure 3 shown, the method includes: Step 310, signal data acquisition and preprocessing.

[0089] Collect fault signal data from various different types of equipment in the substation, and perform denoising, standardization, and normalization processing to improve the data quality and ensure the accuracy of subsequent analysis. For example, filter out invalid data such as equipment in the substation undergoing major overhaul, unaccepted equipment, and frequently occurring fault signals.

[0090] Step 320, performing clustering analysis and identification on the fault signal data.

[0091] Cluster analysis and identification are performed on the fault signal data through a fault diagnosis model to mine the correlation relationships and identify the fault diagnosis results.

[0092] Step 330: Conduct expert verification on the correlation relationships and fault diagnosis results.

[0093] By analyzing the relationship between the signals and the fault types, the experts further sort out and confirm the correlation rules of the signals (the corresponding relationships between the fault signal data, the correlation relationships, and the fault diagnosis results). Based on the input and feedback from the experts, the expert system optimizes the correlation rules and constructs a knowledge base with them as knowledge entries to provide reliable fault signal identification rules.

[0094] Step 340: Expert feedback and self-learning mechanism.

[0095] Based on the verification and optimization of the expert system, the system designs an expert feedback mechanism. Whenever a fault signal and a fault type are identified and confirmed, the experts can correct and supplement the preliminary diagnosis results generated by the algorithm. The feedback information from the experts will be automatically recorded and used for the subsequent learning and adjustment of the system.

[0096] Self-learning mechanism: During the process of continuous increase in signal data, the system can continuously optimize the existing correlation rules through self-learning. Based on online learning and incremental learning algorithms, the system can automatically update the knowledge base and the fault diagnosis model when new signals appear, improving the real-time performance and accuracy of fault diagnosis. Each expert feedback will affect the decision-making process of the system, and the changes made by the experts will cause the system to gradually adjust its algorithm to adapt to new fault types or signal changes. This mechanism enables the system to have a certain adaptive ability and can be dynamically adjusted and optimized as the fault types change. The combination of the expert feedback and the self-learning mechanism ensures that the system can be continuously optimized during long-term operation, improving its ability to identify complex and unknown fault types, and gradually forming a more accurate and reliable fault diagnosis system.

[0097] Step 350: Establishment and update of the knowledge base.

[0098] The knowledge base is continuously updated and improved through the expert system. The signals and correlation rules of each fault event are added to the knowledge base to provide historical data and knowledge support for subsequent fault diagnosis. The knowledge base has an adaptive ability and can continuously optimize and update the correlation rules according to the newly occurred fault situations and signal changes.

[0099] Step 360: System iteration and optimization.

[0100] Based on the continuously updated knowledge base, the system performs iterative optimization through the K-means clustering algorithm to enhance the processing ability of the clustering algorithm for new fault signals. After each iteration, the system can automatically optimize the clustering center and more accurately identify and predict the fault signals of substation equipment.

[0101] Step 370: Fault signal processing and auxiliary decision-making.

[0102] Using the updated association rules in the knowledge base, the system can process the fault signals of substation equipment in real time, identify the fault types, and provide auxiliary decision-making support. The system provides accurate fault warnings for operation and maintenance personnel and recommends the best processing solutions based on historical data and the knowledge base.

[0103] The substation fault diagnosis method provided by the embodiments of the present invention performs clustering analysis and identification on substation fault alarm signals, and excavates the potential correlations between various links when a fault occurs. Through the methods of expert evaluation and iterative optimization, an accurate substation signal association rule base is formed. These rules can effectively guide the automatic monitoring and fault handling of the substation profession, provide intelligent support for operation and maintenance work, and improve the stability and response speed of the system.

[0104] The device provided by the embodiments of the present invention will be described below. The device described below can be mutually corresponding and referred to the method described above.

[0105] Figure 4 is a schematic structural diagram of the substation fault diagnosis device provided by the present invention, as Figure 4 shown. The device includes: An acquisition module 410, configured to acquire the fault signal data of each device in the target substation; A diagnosis module 420, configured to input the fault signal data of each device into the fault diagnosis model to obtain the fault association relationship between the fault signal data output by the fault diagnosis model and the fault diagnosis result of the target substation; A verification module 430, configured to respectively perform expert verification on the fault association relationship and the fault diagnosis result, and generate an association relationship verification result and a fault diagnosis verification result; A storage module 440, configured to store the fault signal data, the association relationship verification result, and the fault diagnosis verification result in the fault diagnosis knowledge base; Wherein, the fault diagnosis model is trained based on the fault diagnosis knowledge base.

[0106] The substation fault diagnosis device provided by the embodiment of the present invention acquires the fault signal data of each device in the target substation; inputs the fault signal data of each device into the fault diagnosis model to obtain the fault correlation relationship between the fault signal data output by the fault diagnosis model and the fault diagnosis result of the target substation; respectively conducts expert verification on the fault correlation relationship and the fault diagnosis result to generate the correlation relationship verification result and the fault diagnosis verification result; stores the fault signal data, the correlation relationship verification result, and the fault diagnosis verification result in the fault diagnosis knowledge base; since the fault diagnosis knowledge base is constructed by collecting the fault signal data, the correlation relationship verification result, and the fault diagnosis verification result, the fault diagnosis model is trained according to the fault diagnosis knowledge base, and the fault diagnosis model deeply analyzes the fault signal data to determine the fault correlation relationship and the fault diagnosis result, which can automatically process a large amount of fault signal data, without relying on manual experience for judgment and analysis, improves the efficiency and accuracy of substation fault diagnosis, and provides intelligent support for the operation and maintenance work of the substation.

[0107] Figure 5 is a schematic structural diagram of the electronic device provided by the present invention, as Figure 5 shown, the electronic device may include: a processor (Processor) 510, a communication interface (Communications Interface) 520, a memory (Memory) 530, and a communication bus (Communications Bus) 540. Among them, the processor 510, the communication interface 520, and the memory 530 complete mutual communication through the communication bus 540. The processor 510 can call the logical commands in the memory 530 to execute the methods described in the above embodiments, for example: acquire the fault signal data of each device in the target substation; input the fault signal data of each device into the fault diagnosis model to obtain the fault correlation relationship between the fault signal data output by the fault diagnosis model and the fault diagnosis result of the target substation; respectively conduct expert verification on the fault correlation relationship and the fault diagnosis result to generate the correlation relationship verification result and the fault diagnosis verification result; store the fault signal data, the correlation relationship verification result, and the fault diagnosis verification result in the fault diagnosis knowledge base; where the fault diagnosis model is trained based on the fault diagnosis knowledge base.

[0108] In addition, when the logical commands in the above-mentioned memory can be implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several commands for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0109] The processor in the electronic device provided by the embodiments of the present invention can call the logical instructions in the memory to implement the above method. The specific implementation manners are the same as those of the foregoing method embodiments, and the same beneficial effects can be achieved, which will not be elaborated herein.

[0110] The embodiments of the present invention also provide a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it is used to execute the methods provided in the above various embodiments.

[0111] The specific implementation manners are the same as those of the foregoing method embodiments, and the same beneficial effects can be achieved, which will not be elaborated herein.

[0112] The embodiments of the present invention provide a computer program product, including a computer program, which when executed by a processor, implements the method as described above.

[0113] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. A person of ordinary skill in the art can understand and implement it without creative labor.

[0114] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the essence of the above technical solution, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0115] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A substation fault diagnosis method, characterized in that, Including: Obtaining the fault signal data of each device in the target substation; Inputting the fault signal data of each device into the fault diagnosis model to obtain the fault association relationship between the fault signal data output by the fault diagnosis model and the fault diagnosis result of the target substation; Respectively conducting expert verification on the fault association relationship and the fault diagnosis result to generate an association relationship verification result and a fault diagnosis verification result; Storing the fault signal data, the association relationship verification result, and the fault diagnosis verification result into the fault diagnosis knowledge base; Wherein, the fault diagnosis model is trained based on the fault diagnosis knowledge base.

2. The substation fault diagnosis method according to claim 1, wherein The fault diagnosis model includes a clustering analysis layer and a result recognition layer connected in sequence.

3. The substation fault diagnosis method according to claim 2, characterized in that The step of inputting the fault signal data of each device into the fault diagnosis model to obtain the fault association relationship between the fault signal data output by the fault diagnosis model and the fault diagnosis result of the target substation includes: Inputting the fault signal data of each device into the clustering analysis layer, and the clustering analysis layer clustering the fault signal data of each device to obtain multiple clusters; each cluster includes at least one fault signal data; Inputting the multiple clusters into the result recognition layer, and the result recognition layer identifying the fault signal data in each cluster to obtain the fault association relationship between the fault signal data in each cluster and the fault diagnosis result of each cluster, and determining the fault diagnosis result of the target substation based on the fault diagnosis results of each cluster.

4. The substation fault diagnosis method according to claim 3, characterized in that, The step of inputting the fault signal data of each device into the clustering analysis layer, and the clustering analysis layer clustering the fault signal data of each device to obtain multiple clusters includes: Determining the initial clustering center in the current iteration process; Determining the distance between each fault signal data and the clustering centers of each cluster; Assigning each fault signal data to the clustering center with the minimum distance to obtain the updated clusters; Determining the updated clustering center based on the mean value of the fault signal data in the updated clusters; In the case that the distance between the updated clustering center and the initial clustering center is greater than the preset distance or the number of iterations is less than the preset number of times, taking the updated clustering center as the initial clustering center in the next iteration process and continuing the iteration; In the case that the distance between the updated clustering center and the initial clustering center is less than the preset distance or the number of iterations is greater than the preset number of times, stopping the iteration.

5. The substation fault diagnosis method according to claim 4, wherein The step of determining the distance between each fault signal data and the clustering centers of each cluster includes: Determining the calculation weights of each data item in each fault signal data; Based on the calculation weights of each data item and the differences between each data item in each fault signal data and each data item in the clustering center, determining the distance between each fault signal data and the clustering centers of each cluster.

6. The substation fault diagnosis method according to any one of claims 1 to 5, characterized in that, The method further includes: Determining that the current fault signal data is different from the historical fault signal data; Storing the current fault signal data, as well as the association relationship verification result and the fault diagnosis verification result corresponding to the current fault signal data, into the fault diagnosis knowledge base; Incrementally train the fault diagnosis model based on the current fault signal data, as well as the correlation check result and the fault diagnosis check result corresponding to the current fault signal data.

7. The substation fault diagnosis method according to claim 1, characterized in that The obtaining of the fault signal data of each device in the target substation includes: Obtain the initial signals of each device sent by the data acquisition and monitoring control system, the fault recorder system, and the centralized control system of the target substation; Extract features from the initial signals to obtain the fault signal data of each device; the fault signal data includes device type, fault occurrence time, fault duration, device operating status, and protection device actions; the device operating status includes current, voltage, and temperature.

8. The substation fault diagnosis method according to claim 7, wherein After the extracting features from the initial signals to obtain the fault signal data of each device, the method further includes: Preprocess the fault signal data of each device; the preprocessing includes at least one of data cleaning, data standardization, and data normalization.

9. A substation fault diagnosis device, characterized in that, It includes: An acquisition module, configured to obtain the fault signal data of each device in the target substation; A diagnosis module, configured to input the fault signal data of each device into a fault diagnosis model to obtain the fault correlation relationship between the fault signal data output by the fault diagnosis model, and the fault diagnosis result of the target substation; A check module, configured to respectively perform expert checks on the fault correlation relationship and the fault diagnosis result to generate a correlation check result and a fault diagnosis check result; A storage module, configured to store the fault signal data, the correlation check result, and the fault diagnosis check result into a fault diagnosis knowledge base; Wherein, the fault diagnosis model is trained based on the fault diagnosis knowledge base.

10. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein, When the processor executes the computer program, it implements the substation fault diagnosis method according to any one of claims 1 to 8.

11. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the substation fault diagnosis method according to any one of claims 1 to 8.

12. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the substation fault diagnosis method according to any one of claims 1 to 8.

Images