Intelligent sensing high-voltage distribution cabinet installation detection system and method

Abstract

The present invention discloses an intelligent sensing high-voltage distribution cabinet installation detection system and method, which relates to the field of distribution cabinets. The system includes: a historical fault information set acquisition module for calling network data based on the model of the distribution cabinet to obtain a historical fault information set; a fault mode analysis module for performing fault mode analysis and constructing K detection trigger trees; an installation detection trigger tree acquisition module for evaluating fault processing priorities and performing trigger node fusion to obtain an installation detection trigger tree; an installation detection model tree construction module for constructing an installation detection model tree; and a fault level detection module for performing intelligent sensor scheduling to collect detection data and synchronously triggering the detection model to receive data and detect faults. The system solves the technical problems of insufficient detection comprehensiveness and accuracy, low efficiency, and insufficient early warning capabilities in existing distribution cabinet installations, achieving the technical effect of improving detection comprehensiveness and accuracy, efficiency, and early warning capabilities.

Description

Technical Field

[0001] The present application relates to the field of power distribution cabinets, and in particular to an intelligent sensing high-voltage power distribution cabinet installation detection system and method. Background Art

[0002] In the power system sector, high-voltage distribution cabinets are critical equipment for power distribution and control, and their installation quality is directly related to the safe and stable operation of the power system. With the development of intelligent technology, how to efficiently and accurately detect potential faults in high-voltage distribution cabinets during installation has become a pressing issue. With the advancement of sensor and data processing technologies, some preliminary intelligent detection methods have begun to be applied. These methods typically monitor the operating status of high-voltage distribution cabinets by installing a certain number of sensors and determine whether faults exist through simple data analysis. However, because sensor deployment locations are often based on empirical judgment, comprehensive coverage of key detection points is difficult. Furthermore, the sensors lack coordination, failing to form an integrated detection system. This results in low detection efficiency and insufficient fault warning capabilities.

[0003] In the current related technologies, the installation inspection of high-voltage distribution cabinets has technical problems such as insufficient comprehensiveness and accuracy, low inspection efficiency and insufficient fault warning capabilities. Summary of the Invention

[0004] The present application provides an intelligent sensing high-voltage distribution cabinet installation detection system and method, which uses network data calls to obtain historical fault information sets, perform fault mode analysis and build a detection trigger tree, and then fuse the trigger nodes to obtain the installation detection trigger tree, and construct an installation detection model tree based on this. During the installation process, the sensor gateway performs intelligent scheduling according to the trigger sequence of the installation detection trigger tree, collects detection data, and the installation detection model tree synchronously triggers the node detection model to receive detection data and perform fault level detection, etc., thereby achieving the technical effect of improving the comprehensiveness and accuracy of detection, improving detection efficiency and fault warning capabilities.

[0005] The present application provides an intelligent sensing high-voltage distribution cabinet installation detection system, including: a historical fault information set acquisition module, which is used to call network data according to the model information of the high-voltage distribution cabinet to obtain a historical fault information set; a fault mode analysis module, which is used to perform fault mode analysis on the historical fault information set and construct K detection trigger trees based on the analysis results; an installation detection trigger tree acquisition module, which is used to evaluate the fault processing priority of the K detection trigger trees and fuse the trigger nodes of the K detection trigger trees according to the evaluation results to obtain an installation detection trigger tree; an installation detection model tree construction module, which is used to construct an installation detection model tree according to the installation detection trigger tree; a fault level detection module, which is used to, during the installation of the high-voltage distribution cabinet, the sensor gateway performs intelligent scheduling of sensors according to the trigger sequence of the installation detection trigger tree to collect detection data, and the installation detection model tree synchronously triggers the node detection model according to the trigger sequence of the installation detection trigger tree to receive detection data and perform fault level detection.

[0006] In a possible implementation, the fault mode analysis module includes: a fault type aggregation unit, used to perform fault type aggregation on the historical fault information set to obtain multiple sample fault types; a sample fault type division unit, used to divide the multiple sample fault types into K groups of sample fault types corresponding to K fault categories based on the fault categories; a historical fault information set decomposition unit, used to decompose the historical fault information set according to the K groups of sample fault types to obtain K groups of sample fault information; an association rule mining unit, used to obtain K fault causal chains by performing association rule mining on the K groups of sample fault information; and a detection trigger tree construction unit, used to construct the K detection trigger trees using the K fault causal chains.

[0007] In a possible implementation, the installation detection model tree construction module includes: a sample fault feature data acquisition unit, which is used to interactively obtain multiple sample fault feature data of the multiple sample fault types; an operator call analysis unit, which is used to perform operator call analysis based on the data features of the multiple sample fault feature data to obtain multiple adaptation operators; a fault identification model parameter optimization unit, which is used to use the multiple adaptation operators to construct multiple fault identification models, and then use the multiple sample fault feature data as training data to optimize the parameters of the multiple fault identification models; a fault detection trigger topology construction unit, which is used to construct a fault detection trigger topology based on the mapping relationship between the multiple sample fault types and the trigger nodes in the installation detection trigger tree; and a model migration unit, which is used to migrate the multiple fault identification models to the fault detection trigger topology to complete the construction of the installation detection model tree.

[0008] In a possible implementation, the fault level detection module includes: a fault detection sensor acquisition unit, used to interactively obtain multiple fault detection sensors of the multiple sample fault types; a sensor storage unit, used to store the multiple fault detection sensors in the sensor gateway; a sensor intelligent scheduling unit, used to, during the installation of the high-voltage distribution cabinet, the sensor gateway performs sensor intelligent scheduling according to the triggering sequence of the installation detection trigger tree to obtain the first scheduling sensor of the first sample detection trigger node; a first fault identification model activation unit, used to activate the first fault identification model in the installation detection model tree according to the first sample detection trigger node; the first fault detection unit, used to start the first scheduling The degree sensor collects detection data of the high-voltage distribution cabinet, and after obtaining the first real-time detection data, inputs the first real-time detection data into the first fault identification model to obtain the first fault detection result; the fault level detection unit is used to, if the first fault detection result is 0, the sensor gateway performs intelligent sensor scheduling to collect detection data according to the triggering sequence of the installation detection trigger tree, and the installation detection model tree synchronously triggers the node detection model to receive detection data and perform fault level detection according to the triggering sequence of the installation detection trigger tree; the installation fault troubleshooting unit is used to, if the first fault detection result is not 0, suspend sensor scheduling, and perform installation fault troubleshooting according to the first sample detection trigger node.

[0009] In a possible implementation, the association rule mining unit includes: a sample fault type permutation and combination subunit, used to permutate and combine H sample fault types in the first group of sample fault types to obtain M groups of sample fault types; a fault association vector configuration subunit, used to configure M groups of fault association vectors for the M groups of sample fault types; a frequent item set mining and confidence evaluation subunit, used to perform frequent item set mining and confidence evaluation on the first group of sample fault information with the M groups of fault association vectors as constraints, and output M causal chain vectors and M causal chain confidence values; a first fault causal chain generation subunit, used to generate the first fault causal chain of the H sample fault types based on the M causal chain vectors and the M causal chain confidence values; and a K fault causal chain acquisition subunit, used to perform association rule mining on the K groups of sample fault information and so on to obtain the K fault causal chains.

[0010] In a possible implementation, the frequent item set mining and confidence evaluation subunit includes: a vector extraction component for extracting M first fault vectors and M second fault vectors from the M groups of fault association vectors; a frequent item set mining component for performing time sequence alignment on the first group of sample fault information, and then using the M first fault vectors and M second fault vectors as constraints to perform frequent item set mining on the first group of sample fault information to obtain M first fault supports, M first fault confidences, M second fault supports, and M second fault confidences. a weighted calculation component for performing weighted calculation on the M first fault supports, the M first fault confidences, the M second fault supports and the M second fault confidences to obtain M first causal confidence values ​​and M second causal confidence values; a causal confidence threshold preset component for presetting a causal confidence threshold; a screening component for traversing the M first causal confidence values ​​and the M second causal confidence values ​​using the causal confidence threshold to screen the M causal chain vectors and the M causal chain confidence values ​​from the M groups of sample fault types.

[0011] In a possible implementation, the installation detection trigger tree acquisition module includes: a first detection trigger tree construction unit, which is used to construct a first detection trigger tree based on the first fault causal chain, wherein the first detection trigger tree includes H sample detection trigger nodes corresponding to the H sample fault types, and the H sample detection trigger nodes are identified by the M causal chain confidence values; a trigger priority sorting unit, which is used to sort the trigger priorities of the H sample detection trigger nodes at the same trigger level according to the M causal chain confidence values ​​to obtain a first trigger level sequence; K trigger level sequence acquisition units, which are used to perform fault processing priority evaluation on the K detection trigger trees and obtain K trigger level sequences by analogy; a fault detection priority acquisition unit, which is used to interactively obtain K fault detection priorities of the K fault categories; a trigger node fusion unit, which is used to locate K groups of connection nodes according to the K trigger level sequences, and perform trigger node fusion on the K detection trigger trees at the K groups of connection nodes based on the K fault detection priorities as constraints to complete the construction of the installation detection trigger tree.

[0012] The present application also provides an intelligent sensing high-voltage distribution cabinet installation detection method, including: calling network data according to the model information of the high-voltage distribution cabinet to obtain a historical fault information set; performing fault mode analysis on the historical fault information set, and constructing K detection trigger trees based on the analysis results; evaluating the fault processing priority of the K detection trigger trees, and performing trigger node fusion on the K detection trigger trees according to the evaluation results to obtain an installation detection trigger tree; constructing an installation detection model tree according to the installation detection trigger tree; during the installation of the high-voltage distribution cabinet, the sensor gateway performs intelligent scheduling of sensors according to the trigger sequence of the installation detection trigger tree to collect detection data, and the installation detection model tree synchronously triggers the node detection model according to the trigger sequence of the installation detection trigger tree to receive detection data and detect fault levels.

[0013] The intelligent sensing high-voltage distribution cabinet installation detection system and method proposed in this application is intended to use a historical fault information set acquisition module to call network data according to the model information of the high-voltage distribution cabinet to obtain a historical fault information set, perform fault mode analysis on the historical fault information set through a fault mode analysis module, and construct K detection trigger trees based on the analysis results, perform fault processing priority evaluation on the K detection trigger trees through an installation detection trigger tree acquisition module, and perform trigger node fusion on the K detection trigger trees according to the evaluation results to obtain an installation detection trigger tree, and construct an installation detection model tree according to the installation detection trigger tree through an installation detection model tree construction module. During the installation of the high-voltage distribution cabinet, the sensor gateway performs intelligent sensor scheduling to collect detection data according to the trigger sequence of the installation detection trigger tree through the fault level detection module, and the installation detection model tree synchronously triggers the node detection model according to the trigger sequence of the installation detection trigger tree to perform detection data reception and fault level detection, thereby achieving the technical effect of improving the comprehensiveness and accuracy of detection, and improving detection efficiency and fault warning capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings of the embodiments of the present invention are briefly introduced below. Flowcharts are used in this application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed in precise order. Instead, various steps may be processed in reverse order or simultaneously as needed. Furthermore, other operations may be added to these processes, or one or more operations may be removed from these processes.

[0015] Figure 1 A schematic diagram of the structure of the intelligent sensing high-voltage distribution cabinet installation detection system provided in an embodiment of the present application.

[0016] Figure 2A flow chart of the intelligent sensing high-voltage distribution cabinet installation detection method provided in an embodiment of the present application.

[0017] Description of the accompanying drawings: historical fault information set acquisition module 10, fault mode analysis module 20, installation detection trigger tree acquisition module 30, installation detection model tree construction module 40, fault level detection module 50. DETAILED DESCRIPTION

[0018] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0019] In order to make the purpose, technical solutions and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings. The described embodiments should not be regarded as limiting this application. All other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0020] In the following description, reference is made to “some embodiments”, which describes a subset of all possible embodiments, but it will be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict, and the terms “first\second” involved are merely used to distinguish similar objects and do not represent a specific ordering of the objects. The terms “including” and “having” and any variations are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or modules that are not clearly listed or that are inherent to these processes, methods, products or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application belongs. The terms used herein are for the purpose of describing the embodiments of this application only.

[0021] The embodiment of the present application provides an intelligent sensing high-voltage distribution cabinet installation detection system, such as Figure 1 As shown, the system includes:

[0022] The historical fault information set acquisition module 10 is used to call network data according to the model information of the high-voltage distribution cabinet to obtain the historical fault information set.

[0023] Specifically, the historical fault information set acquisition module 10 connects to a central database or cloud server, uses the model information of the high-voltage distribution cabinet as a query keyword, and retrieves fault records related to the model from a large amount of historical data, including various faults that occurred in the past for the high-voltage distribution cabinet of this model and related information.

[0024] The failure mode analysis module 20 is configured to perform failure mode analysis on the historical failure information set and construct K detection trigger trees based on the analysis results.

[0025] Specifically, using methods such as data mining, machine learning, or statistical analysis, historical fault information is classified, summarized, and summarized to identify different fault patterns. Based on these fault patterns, K detection trigger trees are constructed. Each tree represents a specific fault detection logic, indicating the conditions and sequence for fault detection. When the specific conditions are met, the corresponding detection action is triggered.

[0026] In one possible implementation, the fault mode analysis module 20 includes: a fault type aggregation unit, configured to perform fault type aggregation on the historical fault information set to obtain a plurality of sample fault types; a sample fault type division unit, configured to divide the plurality of sample fault types into K groups of sample fault types corresponding to K fault categories based on the fault categories; a historical fault information set decomposition unit, configured to decompose the historical fault information set according to the K groups of sample fault types to obtain K groups of sample fault information; an association rule mining unit, configured to obtain K fault causal chains by performing association rule mining on the K groups of sample fault information; and a detection trigger tree construction unit, configured to construct the K detection trigger trees using the K fault causal chains.

[0027] Specifically, the fault type aggregation unit preprocesses the fault records in the historical fault information set, extracting key fault features. Then, it applies a clustering algorithm or classification method to aggregate the fault records into multiple sample fault types with similar properties. The sample fault type classification unit classifies the aggregated sample fault types into corresponding fault categories based on preset fault categories (such as electrical faults, mechanical faults, and environmental faults), forming K groups of sample fault types. The historical fault information set decomposition unit splits the original historical fault information set into K subsets based on the sample fault type classification results, with each subset corresponding to a sample fault type within a fault category. The association rule mining unit uses an association rule mining algorithm to mine association rules on the decomposed K groups of sample fault information, identifying causal relationships or correlations between faults and forming K fault causal chains. A fault causal chain describes the conditions under which a fault occurs and the subsequent faults that may be triggered. The detection trigger tree construction unit uses rule-based or graph-theoretic methods to construct K detection trigger trees based on the mined K fault causal chains. Each detection trigger tree uses a key fault node in the fault causal chain as a trigger point and defines the trigger conditions, detection steps, and fault handling logic. This approach aggregates, divides, and decomposes historical fault information into fault types and mines association rules to construct more accurate and targeted detection trigger trees. These detection trigger trees can reflect the failure modes and characteristics of high-voltage distribution cabinets in actual operation, thereby improving the accuracy and efficiency of fault detection.

[0028] In one possible implementation, the association rule mining unit includes: a sample fault type permutation and combination subunit, used to permutate and combine H sample fault types in the first group of sample fault types to obtain M groups of sample fault types; a fault association vector configuration subunit, used to configure M groups of fault association vectors for the M groups of sample fault types; a frequent item set mining and confidence evaluation subunit, used to perform frequent item set mining and confidence evaluation on the first group of sample fault information with the M groups of fault association vectors as constraints, and output M causal chain vectors and M causal chain confidence values; a first fault causal chain generation subunit, used to generate the first fault causal chain of the H sample fault types based on the M causal chain vectors and the M causal chain confidence values; and a K fault causal chain acquisition subunit, used to perform association rule mining on the K groups of sample fault information and so on to obtain the K fault causal chains.

[0029] Specifically, the sample fault type permutation and combination subunit first selects a first set of sample fault types (belonging to a specific fault category, such as electrical faults). It then permutes and combines these sample fault types to generate all possible fault type combinations. These combinations serve as the basis for subsequent frequent item set mining. For each fault type combination, the fault association vector configuration subunit configures a fault association vector. This vector indicates the presence of each fault type in the combination and the potential associations between them. The frequent item set mining and confidence assessment subunit utilizes association rule mining techniques such as the Apriori algorithm or the FP-Growth algorithm, using the fault association vectors as constraints, to perform frequent item set mining on the first set of sample fault information. The mined frequent item sets represent the frequent co-occurrence patterns between fault types. The frequent item set mining and confidence assessment subunit also calculates the confidence scores of these frequent item sets to assess the reliability of the association rules. Based on the mined frequent item sets and their confidence scores, the first fault causal chain generation subunit generates the first fault causal chain for the first set of sample fault types. This causal chain describes the causal or association relationships between the fault types. K fault causal chain acquisition subunits repeat the above process, performing association rule mining on each of the K groups of sample fault information, ultimately obtaining K fault causal chains. Each fault causal chain corresponds to a specific fault category. This implementation method accurately identifies potential fault modes by mining association rules between fault types, thereby improving fault detection accuracy.

[0030] In a possible implementation, the frequent item set mining and confidence evaluation subunit includes: a vector extraction component for extracting M first fault vectors and M second fault vectors from the M groups of fault association vectors; a frequent item set mining component for performing time sequence alignment on the first group of sample fault information, and then performing frequent item set mining on the first group of sample fault information with the M first fault vectors and M second fault vectors as constraints to obtain M first fault supports, M first fault confidences, M second fault supports, and M second Fault confidence; a weighted calculation component for performing weighted calculation on the M first fault supports, M first fault confidences, M second fault supports and M second fault confidences to obtain M first causal confidence values ​​and M second causal confidence values; a causal confidence threshold preset component for preset a causal confidence threshold; a screening component for using the causal confidence threshold to traverse the M first causal confidence values ​​and the M second causal confidence values ​​to screen the M groups of sample fault types to obtain the M causal chain vectors and M causal chain confidence values.

[0031] Specifically, the vector extraction component extracts M first fault vectors and M second fault vectors from M sets of fault association vectors. These vectors represent potential associations between fault types, with the first fault vector representing the prerequisites for a fault to occur, and the second fault vector representing the consequences or subsequent impacts of a fault. The frequent itemset mining component first performs temporal alignment on the first set of sample fault information to ensure the correct temporal order of the fault events. Then, using the M first fault vectors and M second fault vectors as constraints, it mines frequent itemsets (items that frequently appear in the historical dataset) within the first set of sample fault information. During the mining process, the support and confidence of each fault combination are calculated. The support indicates the frequency of an itemset's occurrence in the dataset, while the confidence is a metric used to assess the strength of the association rule. For example, the first fault vector represents the fault pair A→B, and the second fault vector represents the fault pair B→C. For the fault pair A→B, the support indicates the proportion of fault events in the historical fault information set where both A and B occur. Assume this proportion is 60%, meaning that 60% of the fault events contain both A and B. Confidence indicates the probability that B also occurs in a fault event where A occurs. Assuming this probability is 80%, 80% of all fault events that include A also include B. Similarly, the support and confidence of other fault pairs can be calculated.

[0032] The weighted calculation component performs a weighted calculation (weighted sum or weighted average) on the support and confidence of the mined frequent itemsets to account for the importance of different fault combinations and the strength of their associations. This weighted calculation yields a confidence value for each fault causal chain, which serves as an indicator of its reliability. The causal confidence threshold presetting component presets a causal confidence threshold to filter reliable fault causal chains. Only those with confidence values ​​above the threshold are retained. The screening component uses the preset causal confidence threshold to traverse the confidence values ​​of all mined fault causal chains. Only those with confidence values ​​above the threshold are retained as the final output. This implementation, by mining association rules between fault types and evaluating the confidence of these rules, can more accurately identify potential fault modes, thereby improving fault detection accuracy.

[0033] The installation detection trigger tree acquisition module 30 is configured to evaluate the fault processing priorities of the K detection trigger trees and fuse the trigger nodes of the K detection trigger trees according to the evaluation results to obtain an installation detection trigger tree.

[0034] Specifically, a priority evaluation algorithm is used to evaluate K detection trigger trees, assessing the importance and urgency of different fault handling tasks and determining their processing priorities. Based on the evaluation results, the K detection trigger trees are then node-fused. This means that the K detection trigger trees are connected in order of fault handling priority to generate the installation detection trigger tree.

[0035] In one possible implementation, the installation detection trigger tree acquisition module 30 includes: a first detection trigger tree construction unit, used to construct a first detection trigger tree based on the first fault causal chain, wherein the first detection trigger tree includes H sample detection trigger nodes corresponding to the H sample fault types, and the H sample detection trigger nodes are identified by the M causal chain confidence values; a trigger priority sorting unit, used to sort the trigger priorities of the H sample detection trigger nodes at the same trigger level according to the M causal chain confidence values ​​to obtain a first trigger level sequence; K trigger level sequence acquisition units, used to perform fault processing priority evaluation on the K detection trigger trees and obtain K trigger level sequences by analogy; a fault detection priority acquisition unit, used to interactively obtain K fault detection priorities of the K fault categories; a trigger node fusion unit, used to locate K groups of connection nodes according to the K trigger level sequences, and perform trigger node fusion on the K detection trigger trees at the K groups of connection nodes based on the K fault detection priorities as constraints to complete the construction of the installation detection trigger tree.

[0036] Specifically, the first detection trigger tree construction unit constructs the first detection trigger tree based on the first fault causal chain (generated by the association rule mining unit in the fault mode analysis module 20). This involves first identifying sample fault types within the first fault causal chain and creating a corresponding sample detection trigger node for each fault type. The causal chain confidence value is then used to identify the priority or reliability of these nodes. Within the same trigger hierarchy, the trigger priority sorting unit sorts the sample detection trigger nodes based on the causal chain confidence value, forming a first trigger hierarchy sequence to prioritize nodes with a high probability of failure during the detection process. The K trigger hierarchy sequence acquisition unit repeats the trigger priority sorting steps to generate a corresponding trigger hierarchy sequence for each fault category's detection trigger tree. The fault detection priority acquisition unit interacts with the operator or other system components to obtain the fault detection priorities of the K fault categories, i.e., the relative importance of different fault categories. The trigger node fusion unit locates connecting nodes based on the K trigger hierarchy sequences and fuses the trigger nodes using the fault detection priorities as constraints. This generates a trigger hierarchy sequence for each fault category, representing the detection priorities of different fault types within that category. For example, there's one trigger hierarchy sequence for the electrical fault category and another for the mechanical fault category. Connecting nodes are the connection points between different trigger hierarchy sequences. These connections are determined based on the fault detection priority. For example, the last node in the electrical trigger hierarchy sequence (i.e., the lowest-priority fault type in the electrical fault category) might connect to the first node in the mechanical trigger hierarchy sequence (i.e., the highest-priority fault type in the mechanical fault category). After determining the connecting nodes, the detection trigger trees for different fault categories are fused at these nodes. The fused trigger tree contains detection trigger nodes for all fault categories, arranged in a predetermined order and priority. This allows the system to intelligently schedule sensors during the installation of high-voltage distribution cabinets based on the fused trigger tree and detect faults based on priority. Trigger priority refers to the order in which the system detects different fault types during the installation and testing process, with higher-priority fault types being detected first. A trigger node represents a specific fault type in the detection trigger tree. When triggered, the system performs the corresponding detection operation. This implementation builds and optimizes a detection trigger tree (trigger priority sorting and trigger node fusion), enabling the system to intelligently schedule sensors during installation to achieve maximum fault detection coverage with minimal resource consumption.

[0037] The installation detection model tree construction module 40 is used to construct an installation detection model tree according to the installation detection trigger tree.

[0038] Specifically, based on the structure and logic of the installation detection trigger tree, a corresponding installation detection model tree is constructed. Each node in the model tree corresponds to a specific detection model that can receive data collected by sensors and perform fault detection.

[0039] In one possible implementation, the installation detection model tree construction module 40 includes: a sample fault feature data acquisition unit, which is used to interactively obtain multiple sample fault feature data of the multiple sample fault types; an operator call analysis unit, which is used to perform operator call analysis based on the data features of the multiple sample fault feature data to obtain multiple adaptation operators; a fault identification model parameter optimization unit, which is used to use the multiple adaptation operators to construct multiple fault identification models, and then use the multiple sample fault feature data as training data to optimize the parameters of the multiple fault identification models; a fault detection trigger topology construction unit, which is used to construct a fault detection trigger topology based on the mapping relationship between the multiple sample fault types and the trigger nodes in the installation detection trigger tree; and a model migration unit, which is used to migrate the multiple fault identification models to the fault detection trigger topology to complete the construction of the installation detection model tree.

[0040] Specifically, the sample fault feature data acquisition unit interacts with other modules (such as the historical fault information set acquisition module 10) to collect feature data for multiple sample fault types, such as voltage, current, and temperature at the time of the fault. The operator call analysis unit processes the collected sample fault feature data based on its data characteristics (such as data type, data range, and data distribution). The fault identification model parameter optimization unit uses multiple adaptation operators to construct multiple fault identification models (such as neural networks, support vector machines, and decision trees). It uses the sample fault feature data as training data to optimize the model parameters and optimize model performance by adjusting model parameters (such as the learning rate and number of iterations). The fault detection trigger topology construction unit first determines the trigger node corresponding to each sample fault type. Based on the mapping relationship between the multiple sample fault types and the trigger nodes in the installation detection trigger tree (including the trigger node priority), a fault detection trigger topology is constructed. The model migration unit associates each fault identification model with its corresponding trigger node and then migrates the trained fault identification models to the fault detection trigger topology, completing the construction of the installation detection model tree. This implementation fully utilizes historical fault information to build accurate fault detection models. By intelligently scheduling sensors and synchronously triggering fault detection models, it enables real-time monitoring and fault warning during the installation of high-voltage distribution cabinets.

[0041] The fault level detection module 50 is used to, during the installation of the high-voltage distribution cabinet, enable the sensor gateway to intelligently schedule sensors to collect detection data according to the triggering sequence of the installation detection trigger tree, and the installation detection model tree to synchronously trigger the node detection model to receive detection data and perform fault level detection according to the triggering sequence of the installation detection trigger tree.

[0042] Specifically, during the installation of the high-voltage distribution cabinet, the sensor gateway (responsible for connecting the sensor network and the central control system to realize data collection and transmission) intelligently schedules sensors to collect data according to the triggering sequence of the installation detection trigger tree. The installation detection model tree synchronously triggers the corresponding detection model, receives the data collected by the sensor and performs fault level detection, that is, gradually and deeply detects faults according to the preset hierarchical structure. The embodiment of the present application adopts the method of obtaining a historical fault information set through network data call, performing fault mode analysis and constructing a detection trigger tree, and then fusing the trigger nodes to obtain the installation detection trigger tree, and constructing the installation detection model tree accordingly. During the installation process, the sensor gateway performs intelligent scheduling according to the triggering sequence of the installation detection trigger tree, collects detection data, and the installation detection model tree synchronously triggers the node detection model to receive detection data and perform fault level detection and other technical means, thereby achieving the technical effect of improving the comprehensiveness and accuracy of detection, improving detection efficiency and fault warning capabilities.

[0043] In a possible implementation, the fault level detection module 50 includes: a fault detection sensor acquisition unit, used to interactively obtain multiple fault detection sensors of the multiple sample fault types; a sensor storage unit, used to store the multiple fault detection sensors in the sensor gateway; a sensor intelligent scheduling unit, used to, during the installation of the high-voltage distribution cabinet, the sensor gateway performs sensor intelligent scheduling according to the triggering sequence of the installation detection trigger tree to obtain the first scheduling sensor of the first sample detection trigger node; a first fault identification model activation unit, used to activate the first fault identification model in the installation detection model tree according to the first sample detection trigger node; the first fault detection unit, used to start the first A scheduling sensor collects detection data from the high-voltage distribution cabinet, and after obtaining first real-time detection data, inputs the first real-time detection data into the first fault identification model to obtain a first fault detection result; a fault level detection unit is used to, if the first fault detection result is 0, the sensor gateway performs intelligent sensor scheduling to collect detection data according to the triggering sequence of the installation detection trigger tree, and the installation detection model tree synchronously triggers the node detection model to receive detection data and perform fault level detection according to the triggering sequence of the installation detection trigger tree; an installation fault troubleshooting unit is used to, if the first fault detection result is not 0, suspend sensor scheduling, and perform installation fault troubleshooting according to the first sample detection trigger node.

[0044] Specifically, the fault detection sensor acquisition unit extracts corresponding sensor information from a database based on the preset fault type-sensor mapping. The sensor storage unit stores the acquired fault detection sensor information in the sensor gateway for subsequent intelligent scheduling. The sensor intelligent scheduling unit intelligently schedules sensors for data collection based on the triggering sequence of the installation detection trigger tree. During the installation process, the sensor gateway sequentially activates the corresponding sensors for data collection based on the node order and priority of the installation detection trigger tree. The first fault identification model activation unit (using the first sample detection trigger node as an example) activates the corresponding fault identification model in the installation detection model tree based on the triggering node information. The first fault detection unit (using the first sample detection trigger node as an example) activates the scheduled sensors to collect real-time data from the high-voltage distribution cabinet and inputs the data into the activated fault identification model for detection. The model analyzes and processes the data and outputs the fault detection results. The fault level detection unit determines whether to continue sensor scheduling and fault level detection based on the fault detection results. If the test result is fault-free (i.e., the test result is 0), the next sensor is scheduled for testing according to the order of the installation test trigger tree. If the test result is non-zero (i.e., a fault exists), sensor scheduling is suspended and the system is transferred to the installation fault troubleshooting unit for processing. The installation fault troubleshooting unit determines the possible cause and location of the fault based on the fault detection results and the location of the trigger node, and then generates appropriate measures for troubleshooting and handling, such as replacing damaged components and adjusting parameter settings. This implementation method, by constructing an installation test trigger tree and an installation test model tree, enables the system to intelligently schedule sensors for data collection and activate the corresponding fault identification model for testing based on the order and priority of the trigger nodes, thereby improving the accuracy and efficiency of fault detection.

[0045] In the above, refer to Figure 1 The high-voltage distribution cabinet installation detection system based on intelligent sensing according to the embodiment of the present invention is described in detail. Figure 2 A method for detecting installation of a high-voltage power distribution cabinet using intelligent sensing according to an embodiment of the present invention is described.

[0046] The intelligent sensing high-voltage distribution cabinet installation detection method according to the embodiment of the present invention is used to solve the technical problems of insufficient comprehensiveness and accuracy of detection, low detection efficiency and insufficient fault warning capability in the existing high-voltage distribution cabinet installation detection, so as to achieve the technical effect of improving the comprehensiveness and accuracy of detection, and improving the detection efficiency and fault warning capability.

[0047] The intelligent sensing high-voltage distribution cabinet installation detection method includes: calling network data according to the model information of the high-voltage distribution cabinet to obtain a historical fault information set; performing fault mode analysis on the historical fault information set, and constructing K detection trigger trees based on the analysis results; evaluating the fault processing priority of the K detection trigger trees, and performing trigger node fusion on the K detection trigger trees according to the evaluation results to obtain an installation detection trigger tree; constructing an installation detection model tree according to the installation detection trigger tree; during the installation of the high-voltage distribution cabinet, the sensor gateway performs intelligent scheduling of sensors according to the trigger sequence of the installation detection trigger tree to collect detection data, and the installation detection model tree synchronously triggers the node detection model according to the trigger sequence of the installation detection trigger tree to receive detection data and detect fault levels.

[0048] Among them, performing a fault mode analysis on the historical fault information set and constructing K detection trigger trees based on the analysis results may further include: aggregating fault types on the historical fault information set to obtain multiple sample fault types; dividing the multiple sample fault types into K groups of sample fault types corresponding to K fault categories based on fault categories; decomposing the historical fault information set according to the K groups of sample fault types to obtain K groups of sample fault information; performing association rule mining on the K groups of sample fault information to obtain K fault causal chains; and constructing the K detection trigger trees using the K fault causal chains.

[0049] Among them, constructing an installation detection model tree according to the installation detection trigger tree may further include: interactively obtaining multiple sample fault feature data of the multiple sample fault types; performing operator call analysis based on the data features of the multiple sample fault feature data to obtain multiple adaptation operators; after using the multiple adaptation operators to construct multiple fault identification models, using the multiple sample fault feature data as training data to perform parameter optimization of the multiple fault identification models; constructing a fault detection trigger topology based on the mapping relationship between the multiple sample fault types and the trigger nodes in the installation detection trigger tree; migrating the multiple fault identification models to the fault detection trigger topology to complete the construction of the installation detection model tree.

[0050] Among them, during the installation of the high-voltage distribution cabinet, the sensor gateway performs intelligent scheduling of sensors according to the triggering sequence of the installation detection trigger tree to collect detection data, and the installation detection model tree synchronously triggers the node detection model according to the triggering sequence of the installation detection trigger tree to receive detection data and detect fault levels, which may further include: interactively obtaining multiple fault detection sensors of the multiple sample fault types; storing the multiple fault detection sensors in the sensor gateway; during the installation of the high-voltage distribution cabinet, the sensor gateway performs intelligent scheduling of sensors according to the triggering sequence of the installation detection trigger tree to obtain the first scheduling sensor of the first sample detection trigger node; according to the triggering sequence of the first sample detection trigger node in the The installation detection model tree activates the first fault identification model; the first scheduling sensor is started to collect detection data from the high-voltage distribution cabinet, and after obtaining the first real-time detection data, the first real-time detection data is input into the first fault identification model to obtain a first fault detection result; if the first fault detection result is 0, the sensor gateway performs intelligent sensor scheduling to collect detection data according to the triggering sequence of the installation detection trigger tree, and the installation detection model tree synchronously triggers the node detection model to receive detection data and perform fault level detection according to the triggering sequence of the installation detection trigger tree; if the first fault detection result is not 0, the sensor scheduling is suspended, and the installation fault troubleshooting is performed according to the first sample detection trigger node.

[0051] Among them, obtaining K fault causal chains by performing association rule mining on the K groups of sample fault information can further include: permuting and combining H sample fault types in the first group of sample fault types to obtain M groups of sample fault types; configuring M groups of fault association vectors for the M groups of sample fault types; performing frequent item set mining and confidence evaluation on the first group of sample fault information with the M groups of fault association vectors as constraints, and outputting M causal chain vectors and M causal chain confidence values; generating a first fault causal chain of the H sample fault types based on the M causal chain vectors and the M causal chain confidence values; and so on, obtaining the K fault causal chains by performing association rule mining on the K groups of sample fault information.

[0052] Wherein, taking the M groups of fault association vectors as constraints, performing frequent item set mining and confidence evaluation on the first group of sample fault information, and outputting M causal chain vectors and M causal chain confidence values, may further include: extracting M first fault vectors and M second fault vectors from the M groups of fault association vectors; after performing time sequence alignment on the first group of sample fault information, taking the M first fault vectors and M second fault vectors as constraints, performing frequent item set mining on the first group of sample fault information, and obtaining M first fault support, M second fault confidence values, and M first fault support, M second fault confidence values. The method comprises the following steps: first fault confidences, M second fault supports, and M second fault confidences; performing weighted calculation on the M first fault supports, M first fault confidences, M second fault supports, and M second fault confidences to obtain M first causal confidence values ​​and M second causal confidence values; presetting a causal confidence threshold; and using the causal confidence threshold to traverse the M first causal confidence values ​​and the M second causal confidence values ​​to screen the M groups of sample fault types to obtain the M causal chain vectors and the M causal chain confidence values.

[0053] Among them, performing fault processing priority evaluation on the K detection trigger trees, and performing trigger node fusion on the K detection trigger trees according to the evaluation results to obtain an installation detection trigger tree can further include: constructing a first detection trigger tree based on the first fault causal chain, wherein the first detection trigger tree includes H sample detection trigger nodes corresponding to the H sample fault types, and the H sample detection trigger nodes are identified by the M causal chain confidence values; sorting the trigger priorities of the H sample detection trigger nodes at the same trigger level according to the M causal chain confidence values ​​to obtain a first trigger level sequence; and so on, performing fault processing priority evaluation on the K detection trigger trees to obtain K trigger level sequences; interactively obtaining K fault detection priorities of the K fault categories; locating K groups of connection nodes according to the K trigger level sequences, and using the K fault detection priorities as constraints, performing trigger node fusion on the K detection trigger trees at the K groups of connection nodes to complete the construction of the installation detection trigger tree.

[0054] The intelligent sensing high-voltage distribution cabinet installation detection system provided in the embodiment of the present invention can execute the intelligent sensing high-voltage distribution cabinet installation detection method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0055] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules may be used and run on the user terminal and / or server, and the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other and are not used to limit the scope of protection of the present invention.

[0056] The above specific embodiments do not constitute a limitation to the scope of protection of this application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent replacements and improvements made within the spirit and principles of this application should be included in the scope of protection of this application. In some cases, the actions or steps recorded in this application can be performed in an order different from that in the embodiments and can still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. Intelligent sensing high-voltage distribution cabinet installation detection system, characterized by: The system comprises: A historical fault information set acquisition module is used to call network data based on the model information of the high-voltage distribution cabinet to obtain a historical fault information set; A fault mode analysis module, configured to perform fault mode analysis on the historical fault information set and construct K detection trigger trees based on the analysis results; An installation detection trigger tree acquisition module is used to evaluate the fault processing priority of the K detection trigger trees and fuse the trigger nodes of the K detection trigger trees according to the evaluation results to obtain an installation detection trigger tree; An installation detection model tree construction module is used to construct an installation detection model tree according to the installation detection trigger tree; A fault level detection module is configured to: during the installation of the high-voltage distribution cabinet, enable the sensor gateway to intelligently dispatch sensors to collect detection data according to the triggering sequence of the installation detection trigger tree; and enable the installation detection model tree to synchronously trigger the node detection model to receive detection data and perform fault level detection according to the triggering sequence of the installation detection trigger tree; The fault level detection module includes: a fault detection sensor acquisition unit, configured to interactively acquire a plurality of fault detection sensors of a plurality of sample fault types; a sensor storage unit, configured to store the plurality of fault detection sensors in the sensor gateway; A sensor intelligent scheduling unit, configured to enable the sensor gateway to intelligently schedule sensors according to the triggering sequence of the installation detection trigger tree during the installation of the high-voltage distribution cabinet, and obtain a first scheduling sensor for a first sample detection trigger node; A first fault identification model activation unit, configured to activate a first fault identification model in the installation detection model tree according to the first sample detection trigger node; a first fault detection unit, configured to activate the first dispatching sensor to collect detection data from the high-voltage distribution cabinet, and after obtaining first real-time detection data, input the first real-time detection data into the first fault identification model to obtain a first fault detection result; a fault level detection unit, configured to, if the first fault detection result is 0, cause the sensor gateway to intelligently schedule sensors to collect detection data according to the triggering sequence of the installation detection trigger tree, and cause the installation detection model tree to synchronously trigger the node detection model to receive detection data and perform fault level detection according to the triggering sequence of the installation detection trigger tree; The installation fault troubleshooting unit is configured to suspend sensor scheduling if the first fault detection result is non-zero, and trigger the node to perform installation fault troubleshooting according to the first sample detection.

2. The intelligent sensing high-voltage distribution cabinet installation detection system according to claim 1, characterized in that: The failure mode analysis module includes: a fault type aggregation unit, configured to aggregate the fault types of the historical fault information set to obtain a plurality of sample fault types; a sample fault type classification unit, configured to classify the plurality of sample fault types into K groups of sample fault types corresponding to K fault categories based on the fault categories; a historical fault information set decomposition unit, configured to decompose the historical fault information set according to the K groups of sample fault types to obtain K groups of sample fault information; An association rule mining unit, configured to obtain K fault causal chains by performing association rule mining on the K groups of sample fault information; A detection trigger tree construction unit is configured to construct the K detection trigger trees using the K fault causal chains.

3. The intelligent sensing high-voltage distribution cabinet installation detection system according to claim 2, characterized in that: The installation detection model tree construction module includes: a sample fault characteristic data acquisition unit, configured to interactively acquire a plurality of sample fault characteristic data of the plurality of sample fault types; An operator call analysis unit, configured to perform operator call analysis based on the data features of the plurality of sample fault feature data to obtain a plurality of adaptation operators; a fault identification model parameter optimization unit, configured to optimize the parameters of the multiple fault identification models using the multiple adaptation operators to construct the multiple fault identification models, and then use the multiple sample fault feature data as training data; a fault detection trigger topology construction unit, configured to construct a fault detection trigger topology according to a mapping relationship between the plurality of sample fault types and trigger nodes in the installation detection trigger tree; A model migration unit is used to migrate the multiple fault identification models to the fault detection trigger topology to complete the construction of the installation detection model tree.

4. The intelligent sensing high-voltage distribution cabinet installation detection system according to claim 2, characterized in that: The association rule mining unit includes: A sample fault type permutation and combination subunit is used to permutate and combine H sample fault types in the first group of sample fault types to obtain M groups of sample fault types; a fault association vector configuration subunit, configured to configure M groups of fault association vectors for the M groups of sample fault types; A frequent item set mining and confidence evaluation subunit, configured to perform frequent item set mining and confidence evaluation on the first group of sample fault information using the M groups of fault association vectors as constraints, and output M causal chain vectors and M causal chain confidence values; A first fault causal chain generating subunit is configured to generate a first fault causal chain of the H sample fault types according to the M causal chain vectors and the M causal chain confidence values; The K fault causal chain acquisition subunits are used to obtain the K fault causal chains by analogy by performing association rule mining on the K groups of sample fault information.

5. The intelligent sensing high-voltage distribution cabinet installation detection system according to claim 4, characterized in that: The frequent item set mining and confidence evaluation subunit includes: a vector extraction component, configured to extract M first fault vectors and M second fault vectors from the M groups of fault association vectors; a frequent item set mining component, configured to perform time-series alignment on the first set of sample fault information, and then perform frequent item set mining on the first set of sample fault information using the M first fault vectors and the M second fault vectors as constraints to obtain M first fault supports, M first fault confidences, M second fault supports, and M second fault confidences; a weighted calculation component, configured to perform weighted calculation on the M first fault support degrees, the M first fault confidence degrees, the M second fault support degrees, and the M second fault confidence degrees to obtain M first causal confidence values ​​and M second causal confidence values; A causal confidence threshold preset component, used to preset a causal confidence threshold; A screening component is used to traverse the M first causal confidence values ​​and the M second causal confidence values ​​using the causal confidence threshold to obtain the M causal chain vectors and the M causal chain confidence values ​​from the M groups of sample fault types.

6. The intelligent sensing high-voltage distribution cabinet installation detection system according to claim 4, characterized in that: The installation detection trigger tree acquisition module includes: a first detection trigger tree construction unit, configured to construct a first detection trigger tree based on the first fault causal chain, wherein the first detection trigger tree includes H sample detection trigger nodes corresponding to the H sample fault types, and the H sample detection trigger nodes are identified by the M causal chain confidence values; a trigger priority sorting unit, configured to sort the trigger priorities of the H sample detection trigger nodes at the same trigger level according to the M causal chain confidence values ​​to obtain a first trigger level sequence; K trigger level sequence acquisition units, configured to perform fault processing priority evaluation on the K detection trigger trees and obtain K trigger level sequences; a fault detection priority acquisition unit, configured to interactively obtain K fault detection priorities of the K fault categories; A trigger node fusion unit is used to locate K groups of connection nodes according to the K trigger level sequences, and to fuse the K detection trigger trees at the K groups of connection nodes based on the K fault detection priorities, so as to complete the construction of the installation detection trigger tree.

7. Intelligent sensing high-voltage distribution cabinet installation detection method, characterized in that: The method is implemented by the intelligent sensing high-voltage distribution cabinet installation detection system according to any one of claims 1 to 6, and the method includes: Based on the model information of the high-voltage distribution cabinet, network data is called to obtain a historical fault information set; Performing a fault mode analysis on the historical fault information set, and constructing K detection trigger trees based on the analysis results; Performing fault processing priority evaluation on the K detection trigger trees, and fusing trigger nodes on the K detection trigger trees according to the evaluation results to obtain an installation detection trigger tree; Constructing an installation detection model tree according to the installation detection trigger tree; During the installation of the high-voltage distribution cabinet, the sensor gateway performs intelligent sensor scheduling to collect detection data according to the triggering sequence of the installation detection trigger tree, and the installation detection model tree synchronously triggers the node detection model to receive detection data and detect fault levels according to the triggering sequence of the installation detection trigger tree.

Images