Fault diagnosis method and device, computer device, storage medium and program product

Abstract

The application relates to a fault diagnosis method and device, computer equipment, a storage medium and a program product. Actual operation data of a power device is acquired. According to the similarity between the actual operation data and each historical case corresponding to the power device, a target case matching the actual operation data is searched from each historical case. In the case where the target case is found, an initial diagnosis result is determined according to the target case. The initial diagnosis result is verified by using a preset rule and the actual operation data to obtain a target diagnosis result for the power device. The preset rule includes fault operation data characteristics and a preset diagnosis result. The diagnosis result includes at least one of a fault type, a fault cause and a fault treatment scheme. The above scheme quickly reuses past experience through historical case matching, guarantees diagnosis accuracy by combining preset rule verification, reduces misjudgment and missed judgment caused by manual intervention, and improves the accuracy of power device fault diagnosis.

Description

Technical Field

[0001] This application relates to the field of power equipment operation and maintenance technology, and in particular to a fault diagnosis method, device, computer equipment, storage medium and program product. Background Technology

[0002] With the advancement of smart grid construction, the safe and stable operation of power equipment (such as transformers, switchgear, cables, etc.) is becoming increasingly important. As a core tool for monitoring the status of power equipment and discovering potential faults, the accurate diagnosis of test data from power testing instruments is crucial for the operation and maintenance of power equipment.

[0003] Currently, the diagnosis of power test data mainly relies on manual diagnosis. Operation and maintenance personnel need to combine personal experience with relevant standards to analyze and judge the test data. However, due to insufficient experience, misjudgment or omission is prone to occur, resulting in low diagnostic accuracy. Summary of the Invention

[0004] Therefore, it is necessary to provide a fault diagnosis method, device, computer equipment, storage medium, and program product to address the above-mentioned technical problems and improve the accuracy of fault diagnosis for power equipment.

[0005] Firstly, this application provides a fault diagnosis method, including:

[0006] Obtain actual operating data of power equipment;

[0007] Based on the similarity between the actual operating data and the corresponding historical cases of the power equipment, a target case matching the actual operating data is found from each historical case; wherein, each historical case includes at least equipment identification information, historical operating data, and historical diagnostic results;

[0008] If the target case is found, an initial diagnostic result is determined based on the target case;

[0009] The initial diagnostic results are verified using preset rules and the actual operating data to obtain target diagnostic results for the power equipment; wherein, the preset rules include fault operating data characteristics and preset diagnostic results; the diagnostic results include at least one of fault type, fault cause and fault handling plan.

[0010] In one embodiment, the step of verifying the initial diagnostic result using preset rules and the actual operating data to obtain a target diagnostic result for the power equipment includes:

[0011] If the initial diagnostic result and the actual operating data match a preset rule with a confidence level greater than a first threshold, then a target diagnostic result for the power equipment is determined based on the initial diagnostic result.

[0012] If there is a conflict between the preset rule and the initial diagnostic result and / or the actual operating data, a first verification request is generated, and the target diagnostic result is generated based on the verification result of the verification party on the initial diagnostic result based on the first verification request.

[0013] In one embodiment, when there are multiple initial diagnostic results, determining the target diagnostic result for the power equipment based on the initial diagnostic results includes:

[0014] Obtain the repair rate corresponding to each initial diagnostic result that matches a preset rule with a confidence level greater than the first threshold;

[0015] The initial diagnostic result corresponding to the maximum repair rate is used as the target diagnostic result for the power equipment.

[0016] In one embodiment, the preset rules include standard rules and empirical rules; the method further includes:

[0017] If the target case is not found, the actual operating data will be matched with the standard rules;

[0018] If the actual operating data matches the standard rule, then the preset diagnostic result in the standard rule is used as the candidate diagnostic result;

[0019] If the actual operating data does not match the standard rule, then the actual operating data is matched with the empirical rule, and the preset diagnostic result in the empirical rule with a confidence level greater than the second threshold that matches the initial diagnostic result is used as the candidate diagnostic result;

[0020] The target diagnostic result is determined based on the repair rate corresponding to the candidate diagnostic results.

[0021] In one embodiment, determining the target diagnostic result based on the repair rate corresponding to the candidate diagnostic results includes:

[0022] If the repair rate is greater than the third threshold, then the candidate diagnostic result is taken as the target diagnostic result;

[0023] If the repair rate is less than or equal to the third threshold, a second verification request is generated, and the target diagnostic result is generated based on the verification result of the candidate diagnostic result by the verification party based on the second verification request.

[0024] In one embodiment, the method further includes:

[0025] The power equipment is repaired using the target fault handling scheme from the target diagnostic results.

[0026] If the fault of the power equipment is detected to have been eliminated, new historical cases and / or new preset rules are generated based on the equipment identification information of the power equipment, actual operating data and the target diagnostic results.

[0027] Secondly, this application also provides a fault diagnosis device, comprising:

[0028] The acquisition module is used to acquire the actual operating data of power equipment;

[0029] The search module is used to search for a target case that matches the actual operating data from each historical case based on the similarity between the actual operating data and each historical case corresponding to the power equipment; wherein each historical case includes at least equipment identification information, historical operating data and historical diagnostic results;

[0030] The determination module is used to determine an initial diagnostic result based on the target case when the target case is found;

[0031] The verification module is used to verify the initial diagnostic results using preset rules and the actual operating data to obtain the target diagnostic results for the power equipment; wherein, the preset rules include fault operating data characteristics and preset diagnostic results; the diagnostic results include at least one of fault type, fault cause and fault handling plan.

[0032] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0033] Obtain actual operating data of power equipment;

[0034] Based on the similarity between the actual operating data and the corresponding historical cases of the power equipment, a target case matching the actual operating data is found from each historical case; wherein, each historical case includes at least equipment identification information, historical operating data, and historical diagnostic results;

[0035] If the target case is found, an initial diagnostic result is determined based on the target case;

[0036] The initial diagnostic results are verified using preset rules and the actual operating data to obtain target diagnostic results for the power equipment; wherein, the preset rules include fault operating data characteristics and preset diagnostic results; the diagnostic results include at least one of fault type, fault cause and fault handling plan.

[0037] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0038] Obtain actual operating data of power equipment;

[0039] Based on the similarity between the actual operating data and the corresponding historical cases of the power equipment, a target case matching the actual operating data is found from each historical case; wherein, each historical case includes at least equipment identification information, historical operating data, and historical diagnostic results;

[0040] If the target case is found, an initial diagnostic result is determined based on the target case;

[0041] The initial diagnostic results are verified using preset rules and the actual operating data to obtain target diagnostic results for the power equipment; wherein, the preset rules include fault operating data characteristics and preset diagnostic results; the diagnostic results include at least one of fault type, fault cause and fault handling plan.

[0042] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0043] Obtain actual operating data of power equipment;

[0044] Based on the similarity between the actual operating data and the corresponding historical cases of the power equipment, a target case matching the actual operating data is found from each historical case; wherein, each historical case includes at least equipment identification information, historical operating data, and historical diagnostic results;

[0045] If the target case is found, an initial diagnostic result is determined based on the target case;

[0046] The initial diagnostic results are verified using preset rules and the actual operating data to obtain target diagnostic results for the power equipment; wherein, the preset rules include fault operating data characteristics and preset diagnostic results; the diagnostic results include at least one of fault type, fault cause and fault handling plan.

[0047] The aforementioned fault diagnosis methods, devices, computer equipment, storage media, and program products acquire actual operating data of power equipment; based on the similarity between the actual operating data and corresponding historical cases of the power equipment, they search for target cases that match the actual operating data from each historical case; each historical case includes at least equipment identification information, historical operating data, and historical diagnostic results; when a target case is found, an initial diagnostic result is determined based on the target case; the initial diagnostic result is verified using preset rules and actual operating data to obtain a target diagnostic result for the power equipment; the preset rules include fault operating data characteristics and preset diagnostic results; the diagnostic result includes at least one of fault type, fault cause, and fault handling plan. This scheme, by quickly reusing past experience through historical case matching and ensuring diagnostic accuracy through preset rule verification, reduces misjudgments and omissions caused by human intervention, thus improving the accuracy of power equipment fault diagnosis. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0049] Figure 1 This is a flowchart illustrating a fault diagnosis method in one embodiment;

[0050] Figure 2 This is a structural block diagram of a fault diagnosis device in one embodiment;

[0051] Figure 3 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0052] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0053] The fault diagnosis method provided in this application can be applied to the operation and maintenance of power equipment. This method can be executed by a server or a terminal. The server can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services. The terminal can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, projection devices, etc. Portable wearable devices can include smartwatches, smart bracelets, head-mounted devices, etc. Head-mounted devices can be virtual reality (VR) devices, augmented reality (AR) devices, smart glasses, etc.

[0054] In some optional implementations, embodiments of this application may adopt a three-layer architecture of "cloud-edge-device", with each layer having its own division of labor and working together to improve the overall efficiency and reliability of fault diagnosis.

[0055] The terminal layer is responsible for data acquisition and local protection, is compatible with mainstream power testing instruments, uploads data in real time through multiple protocols, and has local caching and breakpoint resume functions to avoid data loss.

[0056] The edge layer is responsible for data preprocessing and preliminary diagnosis. It optimizes data quality through three levels of data processing (cleaning, standardization, and feature extraction), and then uses a lightweight model to initially identify anomalies, reducing the amount of data transmitted to the cloud. After acquiring actual operational data, an edge layer data preprocessing step is added, explaining the basis for data cleaning (normal data range of the case library, abnormal features of the experience library) and the direction of feature extraction (feature correlation of the case library, key features of the experience library). For example, based on the normal range of 0.3%-0.8% for dielectric loss values ​​of 110kV transformers in the historical case library, the edge layer removes 1.5% of noise values ​​and prioritizes extracting the key feature of dielectric loss value fluctuation amplitude.

[0057] The cloud layer includes servers that act as the intelligent brain, integrating a cloud knowledge base, rule engine, self-learning module, and visualization platform to achieve intelligent diagnosis and global management. The cloud knowledge base includes a historical case library, a standard specification library, and an experience rule library, which can be stored on the cloud server. Specifically, the historical case library includes historical cases, the standard specification library includes standard rules, and the experience rule library includes empirical rules.

[0058] The standard specification library stores domestic and international standards in a three-level classification: voltage level, equipment type, and test item. It uses Natural Language Processing (NLP) technology to structure standard thresholds, supporting rapid retrieval and online access, and includes a standard update reminder function. In the rule matching stage of fault diagnosis, the standard specification library provides the source of standard rules, ensuring their authority. For example, the provisions regarding circuit breaker unbalance rate in Power Industry Standard / Recommended 596 (DL / T596) are structured and used as standard rules for diagnostic matching.

[0059] The experience rule base can automatically extract experience through expert interviews and case studies, transforming it into rule-based expressions and assigning credibility ratings. It supports experience review and prioritization, triggering expert review in case of conflicts. During the data acquisition phase, the experience rule base supplements the collection dimensions and accuracy requirements; during the diagnostic phase, it provides experience rules for matching and verification; and during the closed-loop phase, it guides rule optimization and historical case review. For example, if an expert points out a correlation between insulating oil aging and the growth rate of dielectric loss, this experience can be transformed into a rule and incorporated into the experience rule base for expanding collection items and diagnostic matching.

[0060] For example, a MySQL cluster can be used in the server to store structured data, and a Hadoop Distributed File System (HDFS) can be used to store unstructured data. A three-level permission system and a role-based access control (RBAC) model can be configured to ensure data security and traceability.

[0061] For example, a lightweight model can be deployed at the edge layer. Its core is based on pre-processed, high-quality feature data to quickly match abnormal patterns and pre-screen suspected faulty data, preventing invalid data from being uploaded to the cloud. Its recognition logic relies on a cloud knowledge base, avoids complex deep learning operations, and is well-suited to the limited computing resources at the edge layer.

[0062] The specific implementation process for pre-screening suspected faulty data is as follows:

[0063] Step 1: Receive the output results after the three-level data processing at the edge layer, including cleaned valid data, standardized data in a unified format, and extracted key features (such as the fluctuation range of dielectric loss value, water content in oil, partial discharge, etc.). The feature data needs to be aligned with the historical case features and expert experience features in the cloud knowledge base to ensure consistent recognition logic.

[0064] Step 2, Feature Matching: Compare the key features with the fixed abnormal features in the model and calculate the matching degree.

[0065] Step 3: Threshold determination: Refer to the abnormal thresholds in the empirical rule base (such as partial discharge > 50pC) and the normal characteristic ranges in the historical case base (such as dielectric loss of 110kV transformer 0.3%-0.8%) to perform threshold verification on single features or combinations of features.

[0066] Step 4: Result Classification: If there are no anomalies, meaning all feature data are within the normal range and the match rate with the anomaly pattern is <30%, there is no need to upload to the cloud. If there is a suspected anomaly, meaning a single feature exceeds the threshold, or a combination of features matches the anomaly pattern with a match rate of 30%-80%, it is marked as a suspected anomaly, with feature matching details attached, and pushed to the cloud. If there is a high probability of anomaly, meaning multiple features simultaneously exceed the threshold, or the match rate with the anomaly pattern is ≥80%, it is marked as a high-priority suspected anomaly, uploaded to the cloud first, and an urgent handling alert is provided.

[0067] The lightweight model can be trained in the following ways:

[0068] First, normal and abnormal characteristic data of similar equipment can be extracted from the historical case library. Combined with the explicit abnormal characteristic descriptions in the empirical rule library (such as "fluctuation range of dielectric loss value > 0.1% / h" and "water content in oil > 30ppm"), a lightweight training set can be constructed.

[0069] Furthermore, simplified machine learning models (such as simplified random forests and lightweight support vector machines (SVMs)) can be used to simplify algorithm complexity and the number of parameters, reduce computational overhead, and ensure fast operation at the edge layer.

[0070] Finally, with the goal of binary classification of "normal / abnormal", the model is trained using the training set to solidify the mapping relationship between abnormal features and fault suspicion, such as "the fluctuation range of dielectric loss value > 0.1% / h corresponds to suspected insulation abnormality".

[0071] For example, the model's anomaly feature library is synchronized with the cloud knowledge base in real time. When the cloud updates historical cases and preset rules (such as fault characteristics of new equipment), the edge layer lightweight model supplements the new recognition logic through an incremental update mechanism, without having to retrain the overall model.

[0072] During the model optimization phase, a combination model of random forest and convolutional neural network-long short-term memory (CNN-LSTM) can be used. Parameters are optimized through monthly incremental training, and performance metrics such as diagnostic accuracy, false negative rate, and false positive rate are evaluated. Stable updates are ensured through canary deployment. To avoid catastrophic forgetting, canary deployment first tests 10%-20% of edge nodes, and after 72 hours of verification, it is then rolled out to the whole system.

[0073] The updates to the model and knowledge base have a reverse effect on the data acquisition and edge processing stages. For example, if a new rule shows that water content in oil is a key feature, the terminal layer will add a water content sensor or increase the acquisition frequency; the model will increase the feature weight of the fluctuation range of dielectric loss value, and the edge layer will prioritize the calculation of this indicator.

[0074] In one exemplary embodiment, such as Figure 1 As shown, a fault diagnosis method is provided. Taking the application of this method to a server as an example, the method includes the following steps:

[0075] S101, Obtain actual operating data of power equipment.

[0076] Actual operating data refers to various monitoring data generated during the operation of power equipment, including equipment operating parameters and environmental parameters, such as transformer dielectric loss value, oil moisture content, and operating temperature.

[0077] For example, the terminal layer can be adapted to mainstream power testing instruments, using multiple protocols to acquire data in real time. The terminal layer has local caching and breakpoint resume functions to avoid data loss. During the acquisition process, the basic acquisition dimensions are determined by referring to the historical case library, and the acquisition items are supplemented and expanded by combining the experience rule library. At the same time, the acquisition is performed according to the accuracy calibration basis provided by the case library and the mandatory accuracy standard set by the experience rule library.

[0078] For example, data acquisition for a 110kV transformer includes basic data acquisition dimensions such as dielectric loss value, DC resistance, and ambient temperature. Extended data acquisition items include oil moisture content and dielectric loss value growth rate. The acquisition accuracy requirements are dielectric loss value ±0.01% and DC resistance imbalance rate ±0.1%. The data is uploaded to the server in real time via multiple protocols, and the acquired data is cached locally to prevent network interruption.

[0079] S102, based on the similarity between the actual operating data and the corresponding historical cases of the power equipment, find the target case that matches the actual operating data from each historical case.

[0080] Historical cases refer to complete records of past failures of power equipment, including at least equipment identification information (such as equipment number and model), historical operating data (equipment operating parameters at the time of failure), and historical diagnostic results (failure type, cause, and handling solution); target cases refer to cases selected from historical cases that have the highest similarity to current actual operating data.

[0081] For example, a weighted cosine similarity algorithm can be used to calculate the similarity between actual operational data and each historical case. Historical cases are stored categorized by equipment type, fault type, and test project, and high-accuracy retrieval is achieved through Elasticsearch. When calculating similarity, the weights of key features marked in an empirical rule base are simultaneously referenced to ensure matching accuracy. Furthermore, the historical case with the highest similarity can be used as the target case, or a historical case with a similarity exceeding a preset threshold can be used as the target case. The preset threshold can be set according to actual needs and empirical data; for example, it can be set to 90%.

[0082] For example, the server stores 32 historical cases of aging insulating oil in 110kV transformers. Each case includes data such as equipment number, commissioning time, dielectric loss value at the time of failure, and moisture content in the oil. Using the currently collected actual operating data of a dielectric loss value of 0.92% and a moisture content of 35ppm in the oil, a weighted cosine similarity algorithm is used to find a historical case with a similarity of 92%, which is the target case.

[0083] S103, if a target case is found, determine the initial diagnostic result based on the target case.

[0084] The initial diagnostic results refer to the preliminary diagnostic conclusions derived directly from the target case, including possible fault types, fault causes, and corresponding fault handling solutions.

[0085] For example, historical diagnostic results from the target case can be extracted, and appropriate adjustments can be made based on the differences between the current actual operating data and the historical operating data of the target case to form an initial diagnostic result. If there are multiple highly similar cases, their diagnostic results can be extracted and appropriately adjusted to form an initial diagnostic result.

[0086] For example, if the historical diagnostic results for the target case include the fault type as insulating oil aging, the fault cause as excessive moisture in the oil, and the solution as replacing with qualified insulating oil, then based on the differences between the current actual operating data and the target case data, the initial diagnostic results are determined to be: fault type as insulating oil aging, fault cause as excessive moisture content of 35 ppm in the oil, and solution as replacing with qualified insulating oil and testing the moisture content in the oil.

[0087] S104 uses preset rules and actual operating data to verify the initial diagnostic results and obtain the target diagnostic results for the power equipment.

[0088] The preset rules refer to the basis for verifying the diagnostic results, including fault operation data characteristics and corresponding preset diagnostic results. The diagnostic results include at least one of the following: fault type, fault cause, and fault handling plan. The fault type refers to the specific category of fault occurring in the power equipment, such as aging insulating oil or poor winding contact. The fault cause refers to the specific factors leading to the fault, such as excessive moisture in the oil or manufacturing defects. The fault handling plan refers to the specific operational steps to resolve the fault, such as replacing the insulating oil or adjusting the winding contact position.

[0089] For example, the initial diagnostic results and actual operating data can be matched with preset rules. The preset rules can be sorted by confidence level, with priority given to matching preset rules with high confidence levels. If a high-confidence preset rule is matched, the initial diagnostic result can be used as the target diagnostic result. If no preset rule is matched, the initial diagnostic result needs to be further verified and adjusted to obtain the target diagnostic result. The target diagnostic result is the final diagnostic conclusion determined after verification and can be directly used for fault handling.

[0090] The above embodiments quickly reuse past experience by matching historical cases, and combine preset rules to ensure diagnostic accuracy, reduce misjudgments and omissions caused by manual intervention, and improve the accuracy of power equipment fault diagnosis.

[0091] In some optional implementations, the initial diagnostic results are verified using preset rules and the actual operating data to obtain the target diagnostic results for the power equipment. This can be achieved in the following ways:

[0092] If the initial diagnostic results and actual operating data match a preset rule with a confidence level greater than the first threshold, then the target diagnostic result for the power equipment is determined based on the initial diagnostic results.

[0093] If there is a conflict between the preset rules and the initial diagnostic results and / or the actual operating data, a first verification request is generated, and a target diagnostic result is generated based on the verification result of the verification party on the initial diagnostic results based on the first verification request.

[0094] The first threshold is a preset rule confidence threshold used to filter high-confidence rules, for example, it is set to 80%; the first verification request refers to the information sent to the verification party (such as experts or operation and maintenance personnel) for review when there is a rule conflict, which includes actual operating data, initial diagnostic results, conflicting rules, etc.

[0095] For example, during rule matching, the core features of the initial diagnostic results and the key parameters of the actual operating data can be extracted first and compared with the fault operating data features in the preset rules to determine whether the matching conditions are met. During conflict detection, all preset rules can be traversed. If the fault operating data features of a rule match the actual operating data but the preset diagnostic results are inconsistent with the initial diagnostic results, or if the fault operating data features of a rule are related to the initial diagnostic results but contradict the actual operating data, then a conflict is determined, a first verification request is generated and pushed to the verification terminal, and after receiving the verification results from the verification party, the target diagnostic result is generated.

[0096] In the above embodiments, on the one hand, high-confidence rules are screened through confidence thresholds to ensure the reliability of diagnostic results; on the other hand, manual review is introduced as a fallback in case of conflict to avoid the limitations of a single rule judgment and further improve diagnostic accuracy.

[0097] In some optional implementations, when there are multiple initial diagnostic results, the target diagnostic result for the power equipment can be determined based on the initial diagnostic results in the following ways:

[0098] First, the repair rate corresponding to each initial diagnostic result that matches a preset rule with a confidence level greater than the first threshold can be obtained.

[0099] The repair rate refers to the probability that a fault in a power equipment will be eliminated after adopting a fault handling solution from a certain diagnostic result. It can be obtained by statistically analyzing the implementation effect of the fault handling solution in historical cases.

[0100] For example, historical repair rate data corresponding to each initial diagnosis result that matches the high confidence preset rule can be obtained, and the initial diagnosis result with the highest repair rate can be selected after statistical analysis.

[0101] Therefore, the initial diagnostic results corresponding to the maximum repair rate can be used as the target diagnostic results for power equipment.

[0102] For example, the actual operating data of a transformer matched three target cases, with corresponding initial diagnostic results of insulating oil aging (repair rate 93%), poor winding contact (repair rate 85%), and insulating paper aging (repair rate 88%). All three initial diagnostic results matched the preset rule that the confidence level was greater than the first threshold of 80%. After comparing the repair rates, the initial diagnostic result corresponding to insulating oil aging, which had the highest repair rate, was determined as the target diagnostic result.

[0103] Among the above solutions, the one with the highest repair rate is selected from multiple feasible diagnostic results to improve the success rate of fault handling, reduce the cost waste caused by repeated repairs, and improve operation and maintenance efficiency.

[0104] In some optional implementation methods, the preset rules may include standard rules and empirical rules. Standard rules may be rules formed based on relevant domestic and foreign power equipment standards, while empirical rules may be rules formed based on expert experience and historical case extraction.

[0105] Based on this, if no target case is found, the actual operating data can be matched with standard rules. If the actual operating data matches the standard rules, the preset diagnostic results in the standard rules are used as candidate diagnostic results. If the actual operating data does not match the standard rules, the actual operating data is matched with empirical rules, and the preset diagnostic results in the empirical rules whose confidence level matches the initial diagnostic results is greater than a second threshold are used as candidate diagnostic results. Here, the second threshold is the confidence threshold of the empirical rules, used to filter suitable empirical rules, for example, set to 75%. Candidate diagnostic results refer to diagnostic results obtained through rule matching that require further verification.

[0106] For example, if no target case is found, the actual operating data can be compared with the standard rules one by one. The standard rules are stored in categories according to voltage level, equipment type and test items. The structured processing facilitates quick matching. If there is no match, the data is compared with the empirical rules. The empirical rules with a confidence level greater than the second threshold and matching the initial diagnostic results are selected, and their preset diagnostic results are extracted as candidate diagnostic results.

[0107] Furthermore, the target diagnostic result can be determined based on the repair rate corresponding to the candidate diagnostic results.

[0108] For example, the candidate diagnostic result with the highest repair rate can be used as the target diagnostic result.

[0109] In the above embodiments, when no historical cases are found, fault diagnosis is achieved through hierarchical matching of standard rules and empirical rules, which expands the scope of diagnosis, adapts to new equipment or new fault scenarios, and ensures the continuity and comprehensiveness of diagnosis.

[0110] In some optional implementations, determining the target diagnostic result based on the repair rate corresponding to the candidate diagnostic results can be achieved in the following ways:

[0111] If the repair rate is greater than the third threshold, the candidate diagnostic result is used as the target diagnostic result; if the repair rate is less than or equal to the third threshold, a second verification request is generated, and the target diagnostic result is generated based on the verification result of the candidate diagnostic result by the verification party based on the second verification request.

[0112] The third threshold is the critical value of the repair rate, used to judge the reliability of the candidate diagnostic results, for example, it is set to 80%; the second verification request is the review request sent to the verification party when the repair rate of the candidate diagnostic results does not meet the standard, which includes the candidate diagnostic results, actual running data, rule matching status, etc.

[0113] For example, the historical repair rate corresponding to the candidate diagnostic results can be statistically analyzed. If the repair rate is greater than the third threshold, it is directly determined as the target diagnostic result. If the repair rate does not meet the threshold, a second verification request is generated and submitted to the verification party. The verification party reviews the request based on its professional knowledge and the actual situation, and then provides feedback on the verification result. The target diagnostic result is generated based on the verification result.

[0114] For example, if the third threshold is set to 80%, and a candidate diagnostic result is a partial discharge fault in the current transformer with a corresponding repair rate of 75% (less than the third threshold), a second verification request is generated. After expert review, it is found that although the partial discharge quantity in the actual operating data does not reach the standard threshold, other latent characteristics exist. The corrected diagnostic result is determined to be transformer insulation aging, and this corrected result is ultimately determined as the target diagnostic result. If the repair rate of the candidate diagnostic result is 85%, it is directly used as the target diagnostic result.

[0115] In the above embodiments, high-reliability candidate diagnostic results are screened by a repair rate threshold. If the threshold is not met, manual review is introduced to ensure the effectiveness of the diagnostic results and reduce the risk of fault handling.

[0116] In some optional implementations, new historical cases and new preset rules can be generated based on actual operation and maintenance data to adapt historical cases and preset rules to more operation and maintenance scenarios. This is achieved through three core actions: new case learning, rule iteration, and model optimization. With operation and maintenance feedback and expert corrections as input, the knowledge base and model are updated synchronously, and the preceding steps are ultimately optimized in reverse.

[0117] For example, the target fault handling scheme in the target diagnostic results can be used to repair the power equipment; if the fault of the power equipment is detected to have been eliminated, new historical cases and / or new preset rules are generated based on the equipment identification information of the power equipment, actual operating data and target diagnostic results.

[0118] Among them, the target fault handling plan refers to the fault handling operation steps clearly defined in the target diagnosis results; the new historical cases refer to records containing complete information about the current equipment fault; and the new preset rules refer to rules extracted from the current fault handling process that can be used for subsequent diagnosis.

[0119] For example, power equipment can be repaired according to the target fault handling plan. After repair, the equipment's operating status can be continuously monitored to detect whether the fault has been eliminated. If the fault is eliminated, equipment identification information, actual operating data, and target diagnostic results are collected and organized according to five dimensions (basic information, experimental data, diagnostic results, handling plan, and effect verification). After dual review by the system's initial review and expert final review, the data is confirmed to be complete and free of high duplication, and then categorized and stored in the historical case library. At the same time, new rules are extracted from new cases, and after support and confidence verification, they are included in the preset rule library. For example, the Apriori association rule algorithm can be used to mine associations from new historical cases, calculate support and confidence, and after both thresholds are met, the credibility is labeled by experts and then included in the preset rule library.

[0120] For example, the target fault handling solution for a certain transformer was to replace the insulating oil. After repair, the dielectric loss value was monitored and decreased from 0.92% to 0.7%, and the fault was eliminated. The system collected the transformer's equipment number, commissioning time, actual operating data, target diagnostic results, operating procedures for replacing the insulating oil, and post-repair monitoring data. The system initially reviewed and confirmed that the five dimensions of data were complete and non-duplicative. Experts then conducted a final review to verify that the diagnostic results matched the data and that the handling solution was feasible. This case was then stored as a new historical case under the 110kV transformer insulating oil aging category. The rule "Dielectric loss value 0.85%-0.95% + oil moisture > 30ppm → insulating oil aging" was extracted from this case. The support was calculated to be 6% and the confidence level 88%. After meeting the standard, it was included in the preset rule library.

[0121] In the above embodiments, a self-learning closed loop is formed. As new cases are accumulated, the historical case library and the preset rule library are continuously optimized to improve the system's adaptability to new types of faults, thereby continuously improving the diagnostic accuracy. At the same time, fault handling experience is inherited to reduce subsequent operation and maintenance costs.

[0122] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0123] Based on the same inventive concept, this application also provides a fault diagnosis device for implementing the fault diagnosis method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more fault diagnosis device embodiments provided below can be found in the limitations of the fault diagnosis method described above, and will not be repeated here.

[0124] In one exemplary embodiment, such as Figure 2 As shown, a fault diagnosis device is provided, comprising:

[0125] Module 10 is used to acquire actual operating data of power equipment;

[0126] The search module 20 is used to search for a target case that matches the actual operating data from each historical case based on the similarity between the actual operating data and each historical case corresponding to the power equipment; wherein each historical case includes at least equipment identification information, historical operating data and historical diagnostic results;

[0127] The determination module 30 is used to determine an initial diagnostic result based on the target case when the target case is found;

[0128] The verification module 40 is used to verify the initial diagnostic result using preset rules and the actual operating data to obtain the target diagnostic result for the power equipment; wherein, the preset rules include fault operating data characteristics and preset diagnostic results; the diagnostic result includes at least one of fault type, fault cause and fault handling plan.

[0129] In one embodiment, the verification module 40 specifically includes:

[0130] The determining unit is configured to determine a target diagnostic result for the power equipment based on the initial diagnostic result if the initial diagnostic result and the actual operating data match a preset rule with a confidence level greater than a first threshold.

[0131] The generation unit is configured to generate a first verification request if there is a conflict between a preset rule and the initial diagnostic result and / or the actual operating data, and generate the target diagnostic result based on the verification result of the verification party on the initial diagnostic result based on the first verification request.

[0132] In one embodiment, when there are multiple initial diagnostic results, the determining unit is specifically used for:

[0133] Obtain the repair rate corresponding to each initial diagnostic result that matches a preset rule with a confidence level greater than a first threshold; take the initial diagnostic result corresponding to the maximum repair rate as the target diagnostic result for the power equipment.

[0134] In one embodiment, the preset rules include standard rules and empirical rules; the determining module 30 is further configured to:

[0135] If the target case is not found, the actual operating data is matched with the standard rules; if the actual operating data matches the standard rules, the preset diagnostic results in the standard rules are used as candidate diagnostic results; if the actual operating data does not match the standard rules, the actual operating data is matched with the empirical rules, and the preset diagnostic results in the empirical rules with a confidence level greater than the second threshold that match the initial diagnostic results are used as candidate diagnostic results; the target diagnostic result is determined based on the repair rate corresponding to the candidate diagnostic results.

[0136] In one embodiment, the determining module 30 is specifically used for:

[0137] If the repair rate is greater than the third threshold, the candidate diagnostic result is used as the target diagnostic result; if the repair rate is less than or equal to the third threshold, a second verification request is generated, and the target diagnostic result is generated based on the verification result of the verification party on the candidate diagnostic result based on the second verification request.

[0138] In one embodiment, the device further includes a generation module for:

[0139] The power equipment is repaired using the target fault handling scheme in the target diagnostic results; if the fault of the power equipment is detected to have been eliminated, new historical cases and / or new preset rules are generated based on the equipment identification information of the power equipment, actual operating data and the target diagnostic results.

[0140] Each module in the aforementioned fault diagnosis device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0141] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 3As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores case data and rule data. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When executed by the processor, the computer program implements a fault diagnosis method.

[0142] Those skilled in the art will understand that Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0143] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the fault diagnosis method described in any of the above embodiments.

[0144] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps of the fault diagnosis method described in any of the above embodiments.

[0145] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps of the fault diagnosis method described in any of the above embodiments.

[0146] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0147] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0148] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0149] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A fault diagnosis method, characterized in that, The method includes: Obtain actual operating data of power equipment; Based on the similarity between the actual operating data and the corresponding historical cases of the power equipment, a target case matching the actual operating data is found from each historical case; wherein, each historical case includes at least equipment identification information, historical operating data, and historical diagnostic results; If the target case is found, an initial diagnostic result is determined based on the target case; The initial diagnostic results are verified using preset rules and the actual operating data to obtain target diagnostic results for the power equipment; wherein, the preset rules include fault operating data characteristics and preset diagnostic results; the diagnostic results include at least one of fault type, fault cause and fault handling plan.

2. The method according to claim 1, characterized in that, The process of verifying the initial diagnostic results using preset rules and actual operating data to obtain target diagnostic results for the power equipment includes: If the initial diagnostic result and the actual operating data match a preset rule with a confidence level greater than a first threshold, then a target diagnostic result for the power equipment is determined based on the initial diagnostic result. If there is a conflict between the preset rule and the initial diagnostic result and / or the actual operating data, a first verification request is generated, and the target diagnostic result is generated based on the verification result of the verification party on the initial diagnostic result based on the first verification request.

3. The method according to claim 2, characterized in that, When there are multiple initial diagnostic results, determining the target diagnostic result for the power equipment based on the initial diagnostic results includes: Obtain the repair rate corresponding to each initial diagnostic result that matches a preset rule with a confidence level greater than the first threshold; The initial diagnostic result corresponding to the maximum repair rate is used as the target diagnostic result for the power equipment.

4. The method according to claim 1, characterized in that, The preset rules include standard rules and empirical rules; the method further includes: If the target case is not found, the actual operating data will be matched with the standard rules; If the actual operating data matches the standard rule, then the preset diagnostic result in the standard rule is used as the candidate diagnostic result; If the actual operating data does not match the standard rule, then the actual operating data is matched with the empirical rule, and the preset diagnostic result in the empirical rule with a confidence level greater than the second threshold that matches the initial diagnostic result is used as the candidate diagnostic result; The target diagnostic result is determined based on the repair rate corresponding to the candidate diagnostic results.

5. The method according to claim 4, characterized in that, The step of determining the target diagnostic result based on the repair rate corresponding to the candidate diagnostic results includes: If the repair rate is greater than the third threshold, then the candidate diagnostic result is taken as the target diagnostic result; If the repair rate is less than or equal to the third threshold, a second verification request is generated, and the target diagnostic result is generated based on the verification result of the candidate diagnostic result by the verification party based on the second verification request.

6. The method according to any one of claims 1-5, characterized in that, The method further includes: The power equipment is repaired using the target fault handling scheme from the target diagnostic results. If the fault of the power equipment is detected to have been eliminated, new historical cases and / or new preset rules are generated based on the equipment identification information of the power equipment, actual operating data and the target diagnostic results.

7. A fault diagnosis device, characterized in that, The device includes: The acquisition module is used to acquire the actual operating data of power equipment; The search module is used to search for a target case that matches the actual operating data from each historical case based on the similarity between the actual operating data and each historical case corresponding to the power equipment; wherein each historical case includes at least equipment identification information, historical operating data and historical diagnostic results; The determination module is used to determine an initial diagnostic result based on the target case when the target case is found; The verification module is used to verify the initial diagnostic results using preset rules and the actual operating data to obtain the target diagnostic results for the power equipment; wherein, the preset rules include fault operating data characteristics and preset diagnostic results; the diagnostic results include at least one of fault type, fault cause and fault handling plan.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A 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 steps of the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images