Electrical fault detection method and system applied to power equipment, electronic equipment and readable storage medium thereof

Abstract

The invention discloses an electrical fault detection method and system applied to electrical equipment, electronic equipment and a readable storage medium thereof, and relates to the technical field of electrical detection, and the method comprises the steps: collecting the operation data of the electrical equipment in real time; the waveform deviation degree is calculated, if the waveform deviation degree of any operation data and the normal working condition data is larger than a fault threshold value, it is judged that suspicious operation data exist in the equipment, and multi-dimensional fault diagnosis is triggered; the multi-dimensional fault diagnosis comprises the following steps: generating a candidate fault type set; constructing a fault-oriented directed chain in the candidate fault type set; calculating a posterior index of each candidate fault type in the candidate fault type set by adopting a causal reasoning formula; and comprehensively evaluating the fault risk of the equipment component in the power equipment based on the equipment component main body corresponding to each candidate fault type in combination with the posterior index of each candidate fault type. The method has the advantages that the incidence relation between fault types is deeply mined through a causal reasoning mechanism, and the diagnosis precision and systematicness are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the field of electrical detection technology, and in particular to an electrical fault detection method, system, electronic device and readable storage medium thereof applied to power equipment. Background Art

[0002] With the increasing complexity of power equipment and the diversification of operating environments, traditional fault detection methods face significant challenges. Existing technologies mostly rely on single parameter threshold monitoring or static model analysis, which makes it difficult to adapt to dynamic load changes and multi-source interference environments, resulting in high false alarm rates and insufficient sensitivity. For example, electrical parameter detection based on fixed thresholds cannot effectively distinguish between normal fluctuations and early signs of faults; time domain analysis of mechanical vibration signals lacks in-depth exploration of non-stationary features and easily ignores potential fault correlations. In addition, the coupling effects of environmental factors and equipment status are often simplified, resulting in diagnostic results that deviate from actual operating conditions. These problems make it difficult for existing methods to achieve accurate and timely fault warnings in complex scenarios, restricting the development of intelligent operation and maintenance of power equipment.

[0003] Current fault diagnosis systems have significant shortcomings in multi-dimensional collaborative analysis and causal reasoning. Most systems use an isolated fault feature matching mechanism that does not fully consider the conduction relationship between fault types and the functional associations between equipment components, resulting in fragmented diagnostic results. For example, while traditional fuzzy matching algorithms can generate candidate fault sets, they lack the probabilistic quantification of fault chain reactions, making it difficult to assess the risk of compound faults. At the same time, the application of existing technologies to edge computing nodes is mostly limited to data preprocessing, and does not deeply integrate time-frequency domain feature extraction and dynamic time warping algorithms, resulting in inefficient comparisons of real-time data with historical operating conditions. These limitations result in insufficient fault location accuracy, making it impossible to effectively support risk grading and precise maintenance decisions for equipment components. There is an urgent need to build an intelligent detection system that integrates multi-source parameter analysis and dynamic causal reasoning. Summary of the Invention

[0004] To address the aforementioned technical issues, a method, system, electronic device, and readable storage medium for electrical fault detection in power equipment are provided. These methods address the significant shortcomings of current fault diagnosis systems in multi-dimensional collaborative analysis and causal reasoning. Most systems employ an isolated fault feature matching mechanism, failing to fully consider the transmission relationships between fault types and the functional associations between equipment components, resulting in fragmented diagnostic results.

[0005] In order to achieve the above objects, the technical solution adopted by the present invention is: An electrical fault detection system for electric power equipment, comprising: A data acquisition module is used to collect real-time operating data of power equipment, including electrical parameters, mechanical parameters, and environmental parameters. Electrical parameters include voltage, current harmonics, and partial discharge; mechanical parameters include vibration spectrum and soundprint signals; and environmental parameters include temperature and humidity. An edge computing node, connected to the data acquisition module, for preprocessing the operating data, eliminating noise and synchronizing timestamps; The dynamic comparison module, which includes a historical normal operating condition database and a dynamic time warping algorithm processing unit, is used to compare the waveforms of preprocessed real-time data with the historical normal operating condition database, calculate the waveform deviation, and trigger multi-dimensional fault diagnosis when the deviation exceeds the fault threshold; Multi-dimensional fault diagnosis module, including: A time-frequency feature extraction unit, used to extract time-frequency domain features of suspicious operation data; The fault matching unit is connected to the pre-trained electrical fault diagnosis library to perform fuzzy matching on time-frequency domain features and generate a set of candidate fault types; The causal reasoning unit builds a fault-oriented directed chain of candidate fault types based on the historical fault diagnosis experience library and calculates the posterior indicator of each candidate fault type through the causal reasoning formula; Comprehensive assessment unit, used to map candidate fault types to equipment components, conduct comprehensive fault risk assessment based on a posteriori indicators, and generate a test report containing fault risk, functional area, and location coordinates; The storage module is used to store the historical normal operating condition database, the electrical fault diagnosis library and the historical fault diagnosis experience library.

[0006] An electrical fault detection method for electric power equipment, comprising: Real-time collection of operating data of power equipment, including electrical parameters, mechanical parameters, and environmental parameters. Pre-processing of the data by edge computing nodes to eliminate noise and synchronize timestamps. A dynamic time warping algorithm is used to compare real-time data with a historical database of normal operating conditions to calculate waveform deviation. If the waveform deviation between any operating data and the normal operating data is greater than the fault threshold, the equipment is deemed to have suspicious operating data, triggering multi-dimensional fault diagnosis. Otherwise, no response is taken. The multi-dimensional fault diagnosis includes: Extract the time-frequency domain features of suspicious operating data, perform fuzzy matching with the pre-trained electrical fault diagnosis library, and generate a set of candidate fault types; Based on the historical fault diagnosis experience database of electrical equipment, a fault-oriented directed chain is constructed in the set of candidate fault types; The causal inference formula is used to calculate the posterior index of each candidate fault type in the candidate fault type set; Based on the equipment component body corresponding to each candidate fault type and the a posteriori indicators of each candidate fault type, the failure risk of the equipment components in the power equipment is comprehensively evaluated, and a fault detection report for the power equipment is generated. The fault detection report includes the failure risk of the equipment component, the functional area of ​​the equipment component, and the location coordinates.

[0007] Preferably, the construction of a fault-oriented directed chain in a set of candidate fault types based on a historical fault diagnosis experience database of electrical equipment specifically includes: Based on the historical fault diagnosis experience database of electrical equipment, the causal relationship between each element in the candidate fault type set is determined. If a certain fault type The occurrence of another fault type will lead to If the Fault type A fault-oriented chain; Count all fault-directed chains of all elements in the candidate fault type set, and construct the fault-directed set corresponding to the elements in the candidate fault type set. , , for The corresponding fault-oriented set, for The corresponding end fault type of the jth fault-oriented chain, mi is the total number of corresponding fault-oriented chains, and .

[0008] Preferably, the method of calculating the a posteriori indicator of each candidate fault type in the candidate fault type set by using a causal inference formula specifically includes: Fault-Oriented Sets The prior probability corresponding to each element in Its own prior probability, calculated by the causal inference formula The posterior indicator of The causal reasoning formula is as follows: ,in, for The posterior indicator of for The prior probability of for The prior probability of the kth element in , To include The number of fault-oriented sets for the kth element in .

[0009] Compared with the prior art, the present invention has the following beneficial effects: This method deeply explores the correlation between fault types through the causal reasoning mechanism, significantly improving the diagnostic accuracy and systematicity. The fault-oriented directed chain constructed based on the historical fault diagnosis experience library can quantify the probabilistic dependencies in the fault transmission path and dynamically calculate the posterior indicators of candidate faults through causal formulas. It not only identifies the independent risks of single faults, but also reveals the potential causes and chain effects of concurrent faults. This correlation analysis effectively avoids the misjudgment caused by traditional isolated diagnosis. Especially when dealing with complex fault scenarios, it can accurately locate the root cause component of the fault and predict the failure risk of related components, thereby providing a multi-level decision-making basis for equipment maintenance, significantly reducing the missed detection rate and optimizing the allocation of operation and maintenance resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 This is a flow chart of the electrical fault detection method applied to power equipment proposed in Example 1; Figure 2 This is a flow chart of the method for constructing an electrical fault diagnosis library proposed in Example 2; Figure 3 This is a flow chart of the method for generating a candidate fault type set proposed in Example 3; Figure 4 This is a diagram of the architecture of the electronic equipment in this solution; Figure 5 This is a schematic diagram of the computer-readable storage medium structure in this solution.

[0011] The numbers in the figure are: 500 - electronic device; 501 - bus; 502 - CPU; 503 - ROM; 504 - RAM; 505 - communication port; 506 - input / output component; 507 - hard disk; 508 - user interface; 600 - computer-readable storage medium. DETAILED DESCRIPTION

[0012] The following description is intended to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are merely examples, and those skilled in the art may conceive of other obvious variations.

[0013] Example 1

[0014] Reference Figure 1 As shown, an electrical fault detection method applied to power equipment includes: Real-time collection of operating data of power equipment, including electrical parameters, mechanical parameters, and environmental parameters. Pre-processing of the data by edge computing nodes to eliminate noise and synchronize timestamps. By integrating the collaborative collection of multi-source parameters with real-time preprocessing of edge computing nodes, data transmission delay is effectively reduced and data quality is improved. The synchronized time series data after eliminating noise interference lays the foundation for subsequent precise analysis.

[0015] A dynamic time warping algorithm is used to compare real-time data with a historical database of normal operating conditions to calculate waveform deviation. If the waveform deviation between any operating data and the normal operating data is greater than the fault threshold, the equipment is deemed to have suspicious operating data, triggering multi-dimensional fault diagnosis. Otherwise, no response is taken. The specific comparison process of the dynamic time warping algorithm (DTW) is as follows: Z-score standardization is used to normalize real-time data and historical normal operating condition data respectively; Calculate the Euclidean distance of all point pairs in real-time data and historical normal operating condition data; Find a path with the minimum cumulative distance. The cumulative distance of the path is the waveform deviation between the real-time data and the historical normal working condition data.

[0016] The Dynamic Time Warping (DTW) algorithm flexibly matches waveform shapes across time series, resolving misjudgments caused by load fluctuations or sampling rate differences in traditional fixed-window comparisons and significantly improving the robustness of waveform similarity assessment. A deviation-based threshold triggering mechanism adaptively identifies potential anomalies, avoiding redundant diagnosis of normally fluctuating data and reducing computing resource consumption while maintaining sensitivity.

[0017] The multi-dimensional fault diagnosis includes: Extract the time-frequency domain features of suspicious operating data, perform fuzzy matching with the pre-trained electrical fault diagnosis library, and generate a set of candidate fault types; Time-frequency domain feature extraction combined with a fuzzy matching mechanism can simultaneously capture transient signal mutations and frequency domain energy distribution characteristics, overcoming the potential for single-domain features to miss non-stationary faults. The pre-trained diagnostic library leverages historical fault data to enhance matching generalization. Even for novel fault modes, candidate sets can be generated based on similarity, significantly improving fault type coverage and diagnostic tolerance.

[0018] Based on the historical fault diagnosis experience database of electrical equipment, a fault-oriented directed chain is constructed in the set of candidate fault types; By constructing fault-oriented directed chains driven by historical experience, the system explicitly models the causal relationships between fault types (e.g., insulation aging triggering partial discharge, which in turn leads to a short circuit), linking isolated fault characteristics into a systematic fault evolution network. This correlation quantifies the potential paths of fault chain reactions, providing a logical reasoning framework for complex fault diagnosis and avoiding the fragmented diagnostic conclusions often associated with traditional methods that ignore causal relationships.

[0019] The causal inference formula is used to calculate the posterior index of each candidate fault type in the candidate fault type set; The introduction of causal inference formulas quantifies the global contribution of faults. For example, the posterior indicator of a short-circuit fault depends not only on its own feature matching but also on the probability of upstream insulation degradation faults. This more accurately reflects the actual risk and significantly improves the diagnostic confidence of hidden faults and concurrent faults.

[0020] Based on the equipment component body corresponding to each candidate fault type and the a posteriori indicators of each candidate fault type, the failure risk of the equipment components in the power equipment is comprehensively evaluated, and a fault detection report for the power equipment is generated. The fault detection report includes the failure risk of the equipment component, the functional area of ​​the equipment component, and the location coordinates.

[0021] By integrating component-specific risk mapping with a posteriori indicators, fault location is precisely linked from "type determination" to "physical entity." For example, an insulation fault with a high a posteriori indicator is directly linked to the transformer winding coordinates and its heat dissipation area, simultaneously quantifying the component's overall risk level. This mechanism provides decision-making support for operations and maintenance personnel, combining logical reasoning with spatial location, shortening troubleshooting time and guiding targeted maintenance resource scheduling.

[0022] Example 2: Reference Figure 2 As shown, in this embodiment, the construction process of the electrical fault diagnosis library is as follows: Perform modal decomposition on the operating data when the historical fault occurred, and extract the time-frequency domain features in the operating data when the historical fault occurred as the fault feature vector; Performing modal decomposition on historical fault data and extracting time-frequency domain features can effectively capture transient fault characteristics in non-stationary signals and avoid the loss of feature information caused by traditional single analysis in the time domain or frequency domain.

[0023] Combined with the fault diagnosis results of historical faults, each fault feature vector is manually labeled with the fault type to generate the fault standard feature; The process of manually labeling fault types to generate standard features ensures that the fault feature vector strictly corresponds to the actual fault type, solves the problem of label drift caused by noise interference or fuzzy boundaries in automated labeling, and provides a high-quality data foundation for model training.

[0024] Based on all the standard fault features corresponding to each fault type, non-fault standard features equal in number to the standard fault features are extracted from the historical operation database of the electrical equipment. A machine learning algorithm is then used to train a fault diagnosis model corresponding to the fault type. The fault diagnosis model uses the probability of the fault type as output and the time-frequency domain features of the electrical equipment's operating data as input. By introducing the same number of non-fault features as standard fault features into training, the machine learning algorithm can better distinguish the difference characteristics between normal operating conditions and fault conditions, reduce the risk of model overfitting due to sample imbalance, and enhance the recognition sensitivity of early weak fault signals.

[0025] The fault diagnosis models of all fault types are summarized to obtain the electrical fault diagnosis library.

[0026] The electrical fault diagnosis library is formed by integrating diagnostic models of all fault types. It not only covers the precise matching of single fault types, but also can handle complex fault scenarios through multi-model collaborative reasoning. For example, it can simultaneously identify concurrent anomalies caused by insulation aging and mechanical looseness, thereby greatly improving the robustness and engineering applicability of the diagnostic system.

[0027] Example 3: Reference Figure 3 As shown, based on the second embodiment, this embodiment further proposes extracting the time-frequency domain features of the suspicious operation data, performing fuzzy matching with the pre-trained electrical fault diagnosis library, and generating a set of candidate fault types, specifically including: Substitute the time-frequency domain features of the suspicious operation data into each fault diagnosis model in the electrical fault diagnosis library to obtain the prior probability of each fault type of the power equipment; Filter out the fault types whose prior probability is greater than the fuzzy threshold and summarize them to obtain the candidate fault type set U. ,in, is the fault type whose prior probability is greater than the fuzzy threshold, and n is the total number of fault types whose prior probability is greater than the fuzzy threshold.

[0028] Based on the historical fault diagnosis experience database of electrical equipment, the fault-oriented directed chain in the candidate fault type set is constructed, which specifically includes: Based on the historical fault diagnosis experience database of electrical equipment, the causal relationship between each element in the candidate fault type set is determined. If a certain fault type The occurrence of another fault type will lead to If the Fault type A fault-oriented chain; Count all fault-directed chains of all elements in the candidate fault type set, and construct the fault-directed set corresponding to the elements in the candidate fault type set. , , for The corresponding fault-oriented set, for The corresponding end fault type of the jth fault-oriented chain, mi is the total number of corresponding fault-oriented chains, and .

[0029] This method, by constructing a fault-oriented directed chain, enables explicit modeling and systematic analysis of causal relationships between fault types, thoroughly resolving the problems of misdiagnosis and missed detections caused by isolated judgments in traditional diagnosis. Causal relationship mining based on a historical experience database accurately characterizes fault transmission paths (e.g., insulation degradation → partial discharge → short circuit), linking candidate fault types into a logically rigorous fault evolution network. This correlation not only helps identify the immediate risk of the current fault but also predicts the chain reactions it may trigger. For example, a mechanical loosening fault can be linked to bearing overheating and winding deformation through a directed chain, providing early warning of complex faults. Furthermore, the construction of a fault-oriented set quantifies the probabilistic dependencies between faults, upgrading diagnostic results from "single probability ranking" to "global causal network assessment." This significantly improves the ability to reason about systemic equipment failures, providing a multi-dimensional basis for operation and maintenance strategies, from root cause management to associated prevention and control, and significantly reducing the risk of maintenance delays caused by ignoring causal chains.

[0030] The causal inference formula is used to calculate the posterior index of each candidate fault type in the candidate fault type set, specifically including: Fault-Oriented Sets The prior probability corresponding to each element in Its own prior probability, calculated by the causal inference formula The posterior indicator of The causal reasoning formula is as follows: ,in, for The posterior indicator of for The prior probability of for The prior probability of the kth element in , To include The number of fault-oriented sets for the kth element in .

[0031] The causal reasoning formula uses a similar out-link analysis method to the PageRank algorithm, where for The prior probability of Its own baseline failure risk indicator, Represents The failure probability of the kth element in Since a single fault may be caused by multiple faults, such as a short circuit fault may be triggered by multiple insulation aging events, based on the unclear cause of the fault, The prior probability of the kth element in is the number of its corresponding induced fault type to be recorded as its assigned The posterior correction value of All posterior corrections combined Its own prior probability, to evaluate The posterior indicator of .

[0032] By dynamically integrating the prior probability of a fault and the transmission effects of its associated faults into a causal inference formula, fault risk assessment has evolved from isolated probability calculation to a global causal network. Specifically, the formula incorporates the prior probability of associated faults and their frequency of occurrence in the guidance chain. By weighting and superimposing the contributions of these associated faults, the synergistic effects and chain reaction risks among faults are quantified. For example, when a partial discharge fault has a low prior probability, its presence in multiple guidance chains significantly amplifies its posterior indicator, more accurately reflecting its actual harm as a "hidden cause." This mechanism effectively addresses the bias in single-probability assessments caused by traditional methods that ignore inter-fault dependencies. It is particularly suitable for distinguishing primary and secondary faults in complex fault scenarios, ensuring that diagnostic results are more closely aligned with the actual operating status of the equipment. Furthermore, dynamic weighting based on the number of guidance chains adaptively enhances the impact of frequently associated faults, providing a prioritized basis for O&M decision-making that balances direct and potential derivative risks. For example, root-cause faults with high posterior indicators can be prioritized to avoid subsequent cascading downtime.

[0033] Example 4: This embodiment, based on the electrical fault detection methods applied to power equipment in Embodiments 1 to 3, proposes an electrical fault detection system applied to power equipment, including: A data acquisition module is used to collect real-time operating data of power equipment, including electrical parameters, mechanical parameters, and environmental parameters. Electrical parameters include voltage, current harmonics, and partial discharge; mechanical parameters include vibration spectrum and soundprint signals; and environmental parameters include temperature and humidity. An edge computing node, connected to the data acquisition module, for preprocessing the operating data, eliminating noise and synchronizing timestamps; The dynamic comparison module, which includes a historical normal operating condition database and a dynamic time warping algorithm processing unit, is used to compare the waveforms of preprocessed real-time data with the historical normal operating condition database, calculate the waveform deviation, and trigger multi-dimensional fault diagnosis when the deviation exceeds the fault threshold; Multi-dimensional fault diagnosis module, including: A time-frequency feature extraction unit, used to extract time-frequency domain features of suspicious operation data; The fault matching unit is connected to the pre-trained electrical fault diagnosis library to perform fuzzy matching on time-frequency domain features and generate a set of candidate fault types; The causal reasoning unit builds a fault-oriented directed chain of candidate fault types based on the historical fault diagnosis experience library and calculates the posterior indicator of each candidate fault type through the causal reasoning formula; Comprehensive assessment unit, used to map candidate fault types to equipment components, conduct comprehensive fault risk assessment based on a posteriori indicators, and generate a test report containing fault risk, functional area, and location coordinates; The storage module is used to store the historical normal operating condition database, the electrical fault diagnosis library and the historical fault diagnosis experience library.

[0034] Furthermore, the method according to the embodiment of the present application can also be used with the aid of Figure 4 The electronic device architecture shown in FIG. Figure 4 As shown, the electronic device 500 may include a bus 501, one or more CPUs 502, a read-only memory (ROM) 503, a random access memory (RAM) 504, a communication port 505 connected to a network, an input / output component 506, a hard disk 507, etc. The storage device in the electronic device 500, such as the ROM 503 or the hard disk 507, may store an electrical fault detection method for power equipment provided by the present application. The electronic device 500 may also include a user interface 508. Of course, Figure 4 The architecture shown is only exemplary and can be omitted according to actual needs when implementing different devices. Figure 4 One or more components of an electronic device are shown.

[0035] Figure 5 This is a schematic diagram of the computer-readable storage medium structure provided by an embodiment of the present application. Figure 5 As shown, a computer-readable storage medium 600 according to an embodiment of the present application is shown. Computer-readable instructions are stored on the computer-readable storage medium 600. When the computer-readable instructions are executed by the processor, an electrical fault detection method applied to power equipment according to an embodiment of the present application described with reference to the above figures can be executed. The computer-readable storage medium 600 includes, but is not limited to, for example, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory (cache). Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. In summary, the advantages of the present invention are: deep mining of the correlation between fault types through the causal reasoning mechanism, significantly improving the diagnostic accuracy and systematicity. The fault-oriented directed chain constructed based on the historical fault diagnosis experience library can quantify the probabilistic dependencies in the fault conduction path, and dynamically calculate the posterior indicators of candidate faults through causal formulas, which not only identifies the independent risks of single faults, but also reveals the potential causes and chain effects of concurrent faults. This correlation analysis effectively avoids the misjudgment caused by traditional isolated diagnosis. Especially when dealing with complex fault scenarios, it can accurately locate the root cause component of the fault and predict the failure risk of related components, thereby providing a multi-level decision-making basis for equipment maintenance, significantly reducing the missed detection rate and optimizing the allocation of operation and maintenance resources.

[0036] The above shows and describes the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions merely illustrate the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

[0037] It should be noted that references in the specification to "one embodiment," "an embodiment," "exemplary embodiments," "some embodiments," etc. indicate that the described embodiments may include specific features, structures, or characteristics, but not necessarily every embodiment will include such specific features, structures, or characteristics. Furthermore, when specific features, structures, or characteristics are described in conjunction with an embodiment, it is within the knowledge of persons skilled in the relevant art to implement such features, structures, or characteristics in conjunction with other embodiments (whether or not explicitly described).

[0038] In general, terms can be understood, at least in part, from their use in context. For example, depending at least in part on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in the singular sense, or can be used to describe a combination of features, structures, or characteristics in the plural sense. Additionally, the term "based on" can be understood as not necessarily intended to convey an exclusive set of factors, but can instead, depending at least in part on the context, allow for the presence of other factors that are not necessarily explicitly described.

Claims

1. An electrical fault detection method for power equipment, characterized in that: include: Real-time collection of operating data of power equipment, including electrical parameters, mechanical parameters, and environmental parameters. Pre-processing of the data by edge computing nodes to eliminate noise and synchronize timestamps. A dynamic time warping algorithm is used to compare real-time data with a historical database of normal operating conditions to calculate waveform deviation. If the waveform deviation between any operating data and the normal operating data is greater than the fault threshold, the equipment is deemed to have suspicious operating data, triggering multi-dimensional fault diagnosis. Otherwise, no response is taken. The multi-dimensional fault diagnosis includes: Extract the time-frequency domain features of suspicious operating data, perform fuzzy matching with the pre-trained electrical fault diagnosis library, and generate a set of candidate fault types; Based on the historical fault diagnosis experience database of electrical equipment, a fault-oriented directed chain is constructed in the set of candidate fault types; The causal inference formula is used to calculate the posterior index of each candidate fault type in the candidate fault type set; Based on the equipment component body corresponding to each candidate fault type and the a posteriori indicators of each candidate fault type, the failure risk of the equipment components in the power equipment is comprehensively evaluated, and a fault detection report for the power equipment is generated. The fault detection report includes the failure risk of the equipment component, the functional area of ​​the equipment component, and the location coordinates.

2. The electrical fault detection method for power equipment according to claim 1, characterized in that: The dynamic time warping algorithm is used to compare real-time data with historical normal operating condition data to calculate the waveform deviation, specifically including: Z-score standardization is used to normalize real-time data and historical normal operating condition data respectively; Calculate the Euclidean distance of all point pairs in real-time data and historical normal operating condition data; Find a path with the minimum cumulative distance. The cumulative distance of the path is the waveform deviation between the real-time data and the historical normal working condition data.

3. The electrical fault detection method for power equipment according to claim 2, characterized in that: The construction process of the electrical fault diagnosis library is as follows: Perform modal decomposition on the operating data when the historical fault occurred, and extract the time-frequency domain features in the operating data when the historical fault occurred as the fault feature vector; Combined with the fault diagnosis results of historical faults, each fault feature vector is manually labeled with the fault type to generate the fault standard feature; Based on all the standard fault features corresponding to each fault type, non-fault standard features equal in number to the standard fault features are extracted from the historical operation database of the electrical equipment. A machine learning algorithm is then used to train a fault diagnosis model corresponding to the fault type. The fault diagnosis model uses the probability of the fault type as output and the time-frequency domain features of the electrical equipment's operating data as input. The fault diagnosis models of all fault types are summarized to obtain the electrical fault diagnosis library.

4. The electrical fault detection method for power equipment according to claim 3, characterized in that: The extraction of time-frequency domain features of suspicious operating data and fuzzy matching with a pre-trained electrical fault diagnosis library to generate a set of candidate fault types specifically includes: Substitute the time-frequency domain features of the suspicious operation data into each fault diagnosis model in the electrical fault diagnosis library to obtain the prior probability of each fault type of the power equipment; Filter out the fault types whose prior probability is greater than the fuzzy threshold and summarize them to obtain the candidate fault type set U. ,in, is the fault type whose prior probability is greater than the fuzzy threshold, and n is the total number of fault types whose prior probability is greater than the fuzzy threshold.

5. The electrical fault detection method for power equipment according to claim 4, characterized in that: The method of constructing a fault-oriented directed chain in a set of candidate fault types based on a historical fault diagnosis experience database of electrical equipment specifically includes: Based on the historical fault diagnosis experience database of electrical equipment, the causal relationship between each element in the candidate fault type set is determined. If a certain fault type The occurrence of another fault type will lead to If the Fault type A fault-oriented chain; Count all fault-directed chains of all elements in the candidate fault type set, and construct the fault-directed set corresponding to the elements in the candidate fault type set. , , for The corresponding fault-oriented set, for The corresponding end fault type of the jth fault-oriented chain, mi is the total number of corresponding fault-oriented chains, and .

6. The electrical fault detection method for power equipment according to claim 5, characterized in that: The method of using the causal inference formula to calculate the posterior indicator of each candidate fault type in the candidate fault type set specifically includes: Fault-Oriented Sets The prior probability corresponding to each element in Its own prior probability, calculated by the causal inference formula The posterior indicator of The causal reasoning formula is as follows: ,in, for The posterior indicator of for The prior probability of for The prior probability of the kth element in , To include The number of fault-oriented sets for the kth element in .

7. The electrical fault detection method for power equipment according to claim 6, characterized in that: The comprehensive assessment of the failure risk of the device components in the power equipment based on the device component body corresponding to each candidate fault type and the a posteriori indicator of each candidate fault type specifically includes: Map each element in the candidate fault type set U to its corresponding device component; Sum the posterior indices of all fault types mapped to the equipment components whose prior probabilities are greater than the fuzzy threshold as the comprehensive fault index of the equipment components; The failure risk of equipment components is graded based on their comprehensive failure indicators.

8. An electrical fault detection system for power equipment, characterized in that: The method for detecting an electrical fault in an electric power device according to any one of claims 1 to 7 comprises: A data acquisition module is used to collect real-time operating data of power equipment, including electrical parameters, mechanical parameters and environmental parameters; An edge computing node, connected to the data acquisition module, for preprocessing the operating data, eliminating noise and synchronizing timestamps; The dynamic comparison module, which includes a historical normal operating condition database and a dynamic time warping algorithm processing unit, is used to compare the waveforms of preprocessed real-time data with the historical normal operating condition database, calculate the waveform deviation, and trigger multi-dimensional fault diagnosis when the deviation exceeds the fault threshold; Multi-dimensional fault diagnosis module, including: A time-frequency feature extraction unit, used to extract time-frequency domain features of suspicious operation data; The fault matching unit is connected to the pre-trained electrical fault diagnosis library to perform fuzzy matching on time-frequency domain features and generate a set of candidate fault types; The causal reasoning unit builds a fault-oriented directed chain of candidate fault types based on the historical fault diagnosis experience library and calculates the posterior indicator of each candidate fault type through the causal reasoning formula; Comprehensive assessment unit, used to map candidate fault types to equipment components, conduct comprehensive fault risk assessment based on a posteriori indicators, and generate a test report containing fault risk, functional area, and location coordinates; The storage module is used to store the historical normal operating condition database, the electrical fault diagnosis library and the historical fault diagnosis experience library.

9. An electronic device, characterized in that: include: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute an electrical fault detection method applied to power equipment as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing computer-readable instructions, characterized in that: When the computer-readable instructions are executed by a processor, an electrical fault detection method for power equipment according to any one of claims 1 to 7 is implemented.

Images