Power equipment fault positioning method and system based on dynamic knowledge graph

Abstract

The invention discloses a power equipment fault positioning method and system based on a dynamic knowledge graph, and relates to the technical field of power equipment fault analysis, and the method comprises the steps: determining a power equipment monitoring target, and carrying out the operation data collection of the target power equipment; constructing a large-parameter model and small-model adversarial learning framework, and performing output data processing on large-parameter model and small-model data; performing feature analysis on the target power equipment data through the small model, performing parameter prediction on the target power equipment data in combination with the large parameter model, and determining fault analysis data of the target power equipment; constructing an equipment diagnosis knowledge graph, performing node analysis on the fault analysis data of the target equipment, determining a fault data reasoning path of the target power equipment, and outputting fault data of the target power equipment; according to the invention, the artificial dependence of existing equipment diagnosis is made up; the situation of insufficient utilization of multi-source and multi-mode data is improved, and the detection performance of equipment faults is improved.

Description

Technical Field

[0001] The present invention relates to the field of power equipment fault analysis, and specifically to a power equipment fault location method and system based on a dynamic knowledge graph. Background Art

[0002] The diagnosis of the health status of power equipment is the core of power grid asset management and is crucial for the safe and stable operation of the power grid; according to statistics, the faults of power equipment itself are the main causes of power grid accidents; the diagnosis of the health status of power equipment can be divided into single-state diagnosis and multi-state root cause analysis; among them, single-state diagnosis mainly discriminates according to regulations, using threshold analysis, trend analysis, phase comparison, etc.; multi-state root cause analysis mainly relies on expert experience combined with regulations to judge, and timely and accurately find the root cause of equipment defects; however, its manual analysis has high complexity and great difficulty, resulting in the inability to guarantee the accuracy of the diagnosis results, and the inheritance of high-level expert experience often requires long-term accumulation and is difficult to inherit; therefore, it is urgent to solve the technical defects of relying on manual experience for existing equipment diagnosis, insufficient utilization of multi-source and multi-modal data, and inability to guarantee diagnosis accuracy. Summary of the Invention

[0003] The purpose of the present invention is to provide a power equipment fault location method and system based on a dynamic knowledge graph to solve the problems raised in the prior art.

[0004] To achieve the above purpose, the present invention provides the following technical solutions: A power equipment fault location method based on a dynamic knowledge graph, the method includes the following steps: Determine the monitoring target of the power equipment and collect the operation data of the target power equipment; Construct a large-parameter model and a small-model adversarial learning framework, and perform output data processing on the data of the large-parameter model and the small model; Through the small model, perform feature analysis on the data of the target power equipment, and combine the large-parameter model to perform parameter prediction on the data of the target power equipment to determine the fault analysis data of the target power equipment; Based on the fault analysis data of the target power equipment by the large-parameter model and the small model, construct an equipment diagnosis knowledge graph, perform node analysis on the fault analysis data of the target equipment, determine the fault data inference path of the target power equipment, and output the fault data of the target power equipment.

[0005] Further, obtain the power grid equipment operation node network according to the power grid control port, and determine the number and location data of the target power equipment; where the target power equipment includes transformers, high-voltage circuit breakers, and GIS equipment; Retrieve the operation data packet of the target power equipment through the power grid control port, and construct a periodic target power equipment operation data table; the corresponding target power equipment operation data includes dissolved gas in oil data, transformer frequency response data, transformer mechanical state data, breaker winding current state data, GIS partial discharge data, etc.

[0006] Further, construct a large-parameter model and a small-model adversarial learning framework, plan the data output format by analyzing the types of each target power equipment and the corresponding operation data scenarios, and formulate data conversion rules for the output data of the large-parameter model and the small models for analyzing the operation data of each target power equipment, so as to achieve alignment processing of the output data of the large-parameter model and the small models for analyzing the operation data of each target power equipment; for different power equipment and operation data scenarios, the forms of their output data vary greatly. Unifying the format through conversion rules is conducive to improving the efficiency and accuracy of data processing. Based on the output data of the large-parameter model and the small models for analyzing the operation data of each target power equipment after alignment processing, make a comparison and judgment on the output data manually, determine the difference situation of the output data of the large-parameter model and the small models for analyzing the operation data of each target power equipment, and adjust the model network data training structure or parameters of the large-parameter model based on the difference situation, and perform adversarial learning data update on the large-parameter model; if there is a deviation between the output data of the large-parameter model and the small model for analyzing a certain operation data of a certain power equipment, the model network structure or parameters can be adjusted, such as optimizing the output by adjusting the neuron connection weight values in the neural network, continuously updating the model parameters, and strengthening the learning ability of the model.

[0007] Further, retrieve the periodic target power equipment operation data table of each target power equipment, obtain various types of data of the corresponding target power equipment during the corresponding period, and input the obtained data into the corresponding small models for analyzing the operation data of the target power equipment for feature processing; for the operation data of each time point in the corresponding type of period of the target power equipment, construct a feature vector of the corresponding type of periodic operation data. By integrating the eigenvectors corresponding to various types of operation data within a period of the target power equipment, with the target power equipment as the central node, a unified eigenvector space is constructed to determine the set of eigenvectors of the periodic operation data of each target power equipment; through a large-parameter model, various types of operation data in the next adjacent period of each target power equipment are predicted, and a set of eigenvectors of the predicted operation data in the next adjacent period of each target power equipment is constructed; combining the set of eigenvectors of the periodic operation data of each target power equipment in the current period and the set of eigenvectors of the predicted operation data in the next adjacent period, a fault analysis for each target power equipment is carried out, and the faulty power equipment is determined based on the analysis results; where the fault analysis calculation is ; where Fd(H) is the fault index of the target power equipment with the corresponding number H; V(H) m and V x (H) m are respectively the eigenvectors of the operation data corresponding to the type number m in the set of eigenvectors of the periodic operation data in the current period and the set of eigenvectors of the predicted operation data in the next adjacent period of the target power equipment with the corresponding number H; m is the numbering of the operation data type; according to the fault analysis results, the fault indexes of each target power equipment are judged, and by setting a judgment threshold, a fault prompt is given to the power equipment corresponding to the fault index greater than the threshold.

[0008] Furthermore, based on the large-parameter model combined with the small model for analyzing various types of operation data of the target power equipment, the faulty power equipment is determined from the results of the fault analysis of the periodic operation data of each target power equipment; by constructing an equipment diagnosis knowledge graph, with the faulty equipment as the main node, various types of operation data within the period of the faulty equipment as the sub-nodes of layer A, the historical fault type data of the faulty equipment as the sub-nodes of layer B, and the historical fault cause data corresponding to the historical fault types of the faulty equipment as the sub-nodes of layer C; by analyzing the fault indexes of various types of operation data within the period of the faulty power equipment, the periodic fault operation data of the faulty power equipment is determined; where for the analysis of the fault indexes of the operation data, the operation data of the corresponding type in the current period and the operation data of the corresponding type in the next adjacent period are taken, and eigenvectors are constructed for the data at each time point and a fault index analysis is carried out, and its calculation method and judgment method refer to the fault index calculation method of the above-mentioned power equipment. Based on the determined fault operation data, connect the corresponding fault operation data nodes with the device main node in the device diagnosis knowledge graph to construct a fault data chain for the faulty power equipment; through the analysis of the probability distribution of the corresponding fault types for the determined fault operation data of the faulty power equipment, determine the fault type data of the current fault operation data based on the analysis results; match according to the historical fault cause data corresponding to the fault type data of the corresponding fault operation fault data, and construct a fault operation data inference path for the corresponding fault operation data of the current faulty power equipment according to the matching results; analyze the fault proportion of each fault data corresponding to the fault data inference path respectively, and construct a fault data table for the current faulty power equipment based on the analysis results, and output the priority of each fault data; among them, for the analysis of the probability distribution of the corresponding fault types for the fault operation data, the calculation is ; where L(n) is the fault type distribution function value of the fault operation data corresponding to the type number n; n(k), n x (k) and n(k) j correspond to the data values at each time point k in the current period of the fault operation data with the number n, the data values at each time point k in the next period of the fault operation data with the number n, and the average data values at each time point k in the historical period of the fault operation data with the number n respectively; by querying the fault type corresponding to the interval where the fault type distribution function value of the corresponding fault operation data is located, determine the fault type corresponding to the fault data existing in the faulty power equipment in the current period; by determining the historical influence weight of the corresponding fault type and the fault cause, match each fault cause corresponding to the fault type of the current fault operation data; according to the fault type of the corresponding fault operation data and each matching fault cause data, connect the corresponding nodes respectively in the device diagnosis knowledge graph to obtain the inference path existing in the current fault operation data; based on the fault operation data inference path existing in the current faulty power equipment, analyze the proportion of each inference path of each fault operation data respectively, and the analysis calculation is ; where Y(g q (n)) corresponds to the fault proportion of the inference path number g q of the fault operation data with the number n; Fd(n) is the fault index of the fault operation data corresponding to the number n; L(n) i is the distribution function value of the fault type with the corresponding number i of the fault operation data corresponding to the number n; W q,iThe weight value of the historical fault cause data corresponding to the fault type with the corresponding number i and the corresponding number q; since there may be multiple fault causes for the fault type corresponding to a single fault data, there will be multiple inference paths corresponding to a single fault data when constructing the inference path of the fault operation data. Sort the fault proportions of the inference paths corresponding to the fault operation data of the current faulty power equipment, construct the fault data table of the current faulty power equipment, sort it in descending order according to the fault proportion value of the inference path, and output the fault data table. Alert the corresponding fault operation data, corresponding fault type, and fault cause of the faulty power equipment in the fault data table.

[0009] A power equipment fault location system based on a dynamic knowledge graph, the system includes a power equipment monitoring module, a model construction and learning module, a fault determination module, and a fault analysis module; The power equipment monitoring module determines the monitoring target of the power equipment and collects operation data of the target power equipment; the model construction and learning module constructs a large-parameter model and a small-model adversarial learning framework, and performs output data processing on the data of the large-parameter model and the small model; the fault determination module analyzes the characteristics of the target power equipment data through the small model, combines the large-parameter model to predict the parameters of the target power equipment data, and determines the fault analysis data of the target power equipment; the fault analysis module constructs an equipment diagnosis knowledge graph based on the large-parameter model and the small model for the fault analysis data of the target power equipment, performs node analysis on the fault analysis data of the target equipment, determines the fault data inference path of the target power equipment, and outputs the fault data of the target power equipment.

[0010] Further, the power equipment monitoring module includes a target equipment determination unit and a data collection unit; The target equipment determination unit obtains the power grid equipment operation node network according to the power grid control port, and determines the number and location data of the target power equipment; The data collection unit retrieves the operation data packet of the target power equipment through the power grid control port and constructs a periodic target power equipment operation data table.

[0011] Further, the model construction and learning module includes a model data processing unit and a model parameter adjustment unit; The model data processing unit constructs a large-parameter model and a small-model adversarial learning framework, plans the data output format through the analysis of the types of each target power equipment and the corresponding operation data scenarios, and formulates data conversion rules for the output data of the large-parameter model and the small models for the data analysis of each corresponding target power equipment, so as to realize the alignment processing of the output data of the large-parameter model and the small models for the data analysis of each corresponding target power equipment; The model parameter adjustment unit compares and judges the output data manually based on the large parameter model after alignment processing and the output data of each of the small data analysis models for the corresponding target power equipment, determines the difference between the output data of the large parameter model and the output data of each of the small data analysis models for the corresponding target power equipment, and adjusts the model network data training structure or parameters of the large parameter model based on the difference, and updates the adversarial learning data of the large parameter model.

[0012] Further, the fault determination module includes an operation data feature processing unit and a faulty power equipment determination unit; The operation data feature processing unit retrieves the periodic target power equipment operation data tables of each of the target power equipment, obtains various types of data of the periodic operation of the corresponding target power equipment within the corresponding period, and inputs the obtained data into the corresponding small data analysis models for the target power equipment for feature processing; constructs feature vectors for the operation data of each time point of the corresponding type of period of the target power equipment. The faulty power equipment determination unit integrates the feature vectors corresponding to various types of operation data of each of the target power equipment within the period, constructs a unified feature vector space with the corresponding target power equipment as the central node, and determines the set of feature vectors of the periodic operation data of each of the target power equipment; predicts various types of operation data of each of the target power equipment in the adjacent next period through the large parameter model, and constructs a set of feature vectors of the predicted operation data of each of the target power equipment in the adjacent next period; combines the set of feature vectors of the periodic operation data of each of the target power equipment in the current period and the set of feature vectors of the predicted operation data of the adjacent next period, performs fault analysis on each of the target power equipment, and determines the faulty power equipment based on the analysis results.

[0013] Further, the fault analysis module includes a knowledge graph construction unit and a fault data feedback unit; The knowledge graph construction unit determines the faulty power equipment based on the large parameter model combined with the small data analysis models for various types of operation data of the corresponding target power equipment and the fault analysis results of the periodic operation data of each of the target power equipment; constructs an equipment diagnosis knowledge graph, with the faulty equipment as the main node, various types of operation data of the faulty equipment within the period as the sub-nodes of layer A, the historical fault type data of the faulty equipment as the sub-nodes of layer B, and the historical fault cause data corresponding to the historical fault types of the faulty equipment as the sub-nodes of layer C; determines the periodic fault operation data of the faulty power equipment by analyzing the fault indexes of various types of operation data of the faulty power equipment within the period. The fault data feedback unit connects the corresponding fault operation data node with the device main node in the device diagnosis knowledge graph based on the determined fault operation data to construct a fault data chain of the faulty power equipment; analyzes the probability distribution of the corresponding fault types for the determined fault operation data of the faulty power equipment, and determines the fault type data of the current fault operation data based on the analysis results; matches the historical fault cause data corresponding to the fault type data of the corresponding fault operation fault data, and constructs a fault operation data inference path for the corresponding fault operation data of the current faulty power equipment according to the matching results; analyzes the fault proportions of the fault data inference paths corresponding to each fault data respectively, constructs a fault data table of the current faulty power equipment based on the analysis results respectively, and outputs the priorities of each fault data.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows: The present invention performs multi-source data integration processing on monitoring devices by constructing a large-parameter model and a distributed small model, and determines faulty power equipment by performing feature processing on the operation data of power equipment; analyzes the fault data based on the faulty power equipment, and constructs a device diagnosis knowledge graph to perform fault paths on the fault data; the present invention performs adversarial learning on the large-parameter model and the small model, performs feature processing on the device operation data and conducts fault analysis, and then after determining the fault data of the faulty device, analyzes the fault type and fault cause hierarchically, and conducts root cause analysis based on the knowledge graph to construct a fault inference path, so as to realize the output of faulty devices, fault data, fault types and fault causes; the present invention makes up for the artificial dependence of existing device diagnosis; improves the insufficient utilization of multi-source and multi-modal data, and improves the detection performance of device faults. Description of the Drawings

[0015] Figure 1 It is a schematic structural diagram of a power equipment fault location system based on a dynamic knowledge graph of the present invention; Figure 2 It is a schematic flow diagram of a power equipment fault location method based on a dynamic knowledge graph of the present invention. Detailed Embodiments

[0016] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0017] Embodiment: As Figure 1 shown, the present invention provides a technical solution: A power equipment fault location system based on a dynamic knowledge graph, comprising a power equipment monitoring module, a model construction and learning module, a fault determination module, and a fault analysis module; Among them, the power equipment monitoring module determines the monitoring target of the power equipment and collects operation data of the target power equipment; the model construction and learning module constructs a large-parameter model and a small-model adversarial learning framework, and performs output data processing on the data of the large-parameter model and the small model; the fault determination module analyzes the characteristics of the target power equipment data through the small model, combines the large-parameter model to predict the parameters of the target power equipment data, and determines the fault analysis data of the target power equipment; the fault analysis module constructs an equipment diagnosis knowledge graph based on the fault analysis data of the target power equipment by the large-parameter model and the small model, analyzes the nodes of the fault analysis data of the target equipment, determines the inference path of the target power equipment fault data, and outputs the target power equipment fault data.

[0018] Furthermore, the power equipment monitoring module includes a target equipment determination unit and a data collection unit; The target equipment determination unit obtains the grid equipment operation node network according to the grid control port, and determines the number and location data of the target power equipment; The data collection unit retrieves the operation data packet of the target power equipment through the grid control port and constructs a periodic target power equipment operation data table.

[0019] Furthermore, the model construction and learning module includes a model data processing unit and a model parameter adjustment unit; The model data processing unit constructs a large-parameter model and a small-model adversarial learning framework, plans the data output format by analyzing the types of each target power equipment and the corresponding operation data analysis scenarios, and formulates data conversion rules for the output data of the large-parameter model and the small models corresponding to the data analysis of each target power equipment, so as to realize the alignment processing of the output data of the large-parameter model and the small models corresponding to the data analysis of each target power equipment; The model parameter adjustment unit, based on the output data of the large-parameter model and the small models corresponding to the data analysis of each target power equipment after alignment processing, makes a comparison and judgment on the output data manually, determines the difference situation of the output data of the large-parameter model and the small models corresponding to the data analysis of each target power equipment, and adjusts the model network data training structure or parameters of the large-parameter model based on the difference situation, and updates the adversarial learning data of the large-parameter model.

[0020] Furthermore, the fault determination module includes an operation data feature processing unit and a faulty power equipment determination unit; The operation data feature processing unit retrieves the periodic target power equipment operation data tables of each target power equipment, obtains various types of data of the corresponding target power equipment during the corresponding period, and inputs the obtained data into the corresponding target power equipment data analysis sub-model for feature processing; for the operation data of each time point of the corresponding type of cycle of the target power equipment, a feature vector of the corresponding type of cycle operation data is constructed; The faulty power equipment determination unit integrates the feature vectors corresponding to various types of operation data of the corresponding target power equipment during the period, constructs a unified feature vector space with the corresponding target power equipment as the central node, and determines the set of feature vectors of the periodic operation data of each target power equipment; predicts various types of operation data of each target power equipment in the adjacent next period through a large-parameter model, and constructs a set of feature vectors of the predicted operation data of each target power equipment in the adjacent next period; combines the set of feature vectors of the periodic operation data of each target power equipment in the current period and the set of feature vectors of the predicted operation data of the adjacent next period, conducts fault analysis on each corresponding target power equipment, and determines the faulty power equipment based on the analysis results.

[0021] Furthermore, the fault analysis module includes a knowledge graph construction unit and a fault data feedback unit; The knowledge graph construction unit determines the faulty power equipment based on the fault analysis results of the periodic operation data of each target power equipment by combining the large-parameter model with the corresponding target power equipment data analysis sub-models of various types of operation data; constructs an equipment diagnosis knowledge graph, with the faulty equipment as the main node, various types of operation data of the faulty equipment during the period as the A-layer sub-nodes, the historical fault type data of the faulty equipment as the B-layer sub-nodes, and the historical fault cause data corresponding to the historical fault types of the faulty equipment as the C-layer sub-nodes; determines the periodic fault operation data of the faulty power equipment by analyzing the fault indices of various types of operation data of the faulty power equipment during the period. The fault data feedback unit connects the corresponding fault operation data nodes with the equipment main node in the equipment diagnosis knowledge graph based on the determined fault operation data, and constructs a fault data chain of the faulty power equipment; conducts a probability distribution analysis of the corresponding fault types on the determined fault operation data of the faulty power equipment, and determines the fault type data of the current fault operation data based on the analysis results; matches the historical fault cause data corresponding to the fault type data of the corresponding fault operation fault data, and constructs a fault operation data inference path of the current faulty power equipment corresponding to the fault operation data according to the matching results; analyzes the fault proportions of each fault data corresponding to the fault data inference path respectively, constructs a fault data table of the current faulty power equipment based on the analysis results, and outputs the priorities of each fault data. As Figure 2 shown, the present invention provides another technical solution: A power equipment fault location method based on a dynamic knowledge graph, the method comprising the following steps: Determine the monitoring target of the power equipment, and collect the operation data of the target power equipment; Construct a large-parameter model and a small-model adversarial learning framework, and perform output data processing on the data of the large-parameter model and the small model; Analyze the characteristics of the target power equipment data through the small model, combine the large-parameter model to predict the parameters of the target power equipment data, and determine the fault analysis data of the target power equipment; Based on the fault analysis data of the target power equipment by the large-parameter model and the small model, construct an equipment diagnosis knowledge graph, perform node analysis on the fault analysis data of the target equipment, determine the fault data inference path of the target power equipment, and output the fault data of the target power equipment.

[0022] Further, obtain the grid equipment operation node network according to the grid control port, and determine the number and location data of the target power equipment; wherein the target power equipment includes transformers, high-voltage circuit breakers, and GIS equipment; Retrieve the operation data packet of the target power equipment through the grid control port, and construct a periodic target power equipment operation data table; wherein the operation data corresponding to the target power equipment includes dissolved gas in oil data, transformer frequency response data, transformer mechanical state data, breaker winding current state data, and GIS partial discharge data, etc.

[0023] Further, construct a large-parameter model and a small-model adversarial learning framework, plan the data output format through the analysis of the types of each target power equipment and the corresponding operation data analysis scenarios, and formulate data conversion rules for the output data of the large-parameter model and the small models corresponding to the data analysis of each target power equipment, so as to achieve alignment processing of the output data of the large-parameter model and the small models corresponding to the data analysis of each target power equipment; wherein for different power equipment and operation data scenarios, the output data forms are quite different, and the format is unified through conversion rules, which is conducive to improving the efficiency and accuracy of data processing; Based on the output data of the large-parameter model and the small models corresponding to the data analysis of each target power equipment after alignment processing, make a comparison and judgment on the output data manually, determine the difference situation of the output data of the large-parameter model and the small models corresponding to the data analysis of each target power equipment, and adjust the model network data training structure or parameters of the large-parameter model based on the difference situation, and update the adversarial learning data of the large-parameter model; wherein if there is a deviation between the output data of the large-parameter model and the small model corresponding to a certain operation data of a certain power equipment, the model network structure or parameters can be adjusted, such as optimizing the output by adjusting the neuron connection weight value in the neural network, and continuously updating the model parameters to strengthen the learning ability of the model.

[0024] Further, by retrieving the periodic target power equipment operation data tables of each target power equipment, various types of data on the periodic operation of the corresponding target power equipment within the corresponding period are obtained, and the obtained data are respectively input into the corresponding target power equipment data analysis sub-models for feature processing; for the operation data of each time point in the corresponding type period of the target power equipment, a feature vector of the corresponding type period operation data is constructed; By integrating the feature vectors corresponding to various types of operation data within the period of the corresponding target power equipment, with the corresponding target power equipment as the central node, a unified feature vector space is constructed to determine the set of feature vectors of the periodic operation data of each target power equipment; the large-parameter model is used to predict various types of operation data within the adjacent next period of each target power equipment, and a set of feature vectors of the predicted operation data of the adjacent next period of each target power equipment is constructed; combining the set of feature vectors of the periodic operation data of each target power equipment within the current period and the set of feature vectors of the predicted operation data of the adjacent next period, a fault analysis of each corresponding target power equipment is performed, and the faulty power equipment is determined based on the analysis results; among them, the fault analysis calculation is ; where Fd(H) is the fault index of the target power equipment with the corresponding number H; V(H) m and V x (H) m are respectively the feature vectors of the operation data of the corresponding type number m in the set of feature vectors of the periodic operation data of the current period and the set of feature vectors of the predicted operation data of the adjacent next period of the target power equipment with the number H; m is the numbering of the operation data type; according to the fault analysis results, the fault indexes of each target power equipment are judged, and by setting a judgment threshold, a fault prompt is given to the power equipment corresponding to the fault index greater than the threshold.

[0025] Further, based on the large-parameter model combined with the corresponding target power equipment various types of operation data analysis sub-models, the fault analysis results of the periodic operation data of each target power equipment are used to determine the faulty power equipment; by constructing an equipment diagnosis knowledge graph, with the faulty equipment as the main node, various types of operation data within the period of the faulty equipment as the A-layer sub-nodes, the historical fault type data of the faulty equipment as the B-layer sub-nodes, and the historical fault cause data corresponding to the historical fault types of the faulty equipment as the C-layer sub-nodes; by analyzing the fault indexes of various types of operation data within the period of the faulty power equipment, the periodic fault operation data of the faulty power equipment are determined; among them, for the fault index analysis of the operation data, the corresponding type of operation data within the current period and the corresponding type of operation data in the adjacent next period are used, and the vector construction is performed for the data at each time point and the fault index analysis is carried out. The calculation method and judgment method refer to the above-mentioned fault index calculation method of the power equipment; Based on the determined fault operation data, connect the corresponding fault operation data nodes with the device main node in the device diagnosis knowledge graph to construct a fault data chain for the faulty power equipment; through the analysis of the probability distribution of the corresponding fault types for the determined fault operation data of the faulty power equipment, determine the fault type data of the current fault operation data based on the analysis results; match according to the historical fault cause data corresponding to the fault type data of the corresponding fault operation fault data, and construct a fault operation data inference path for the corresponding fault operation data of the current faulty power equipment according to the matching results; analyze the fault proportion of each fault data corresponding to the fault data inference path respectively, and construct a fault data table for the current faulty power equipment based on the analysis results, and output the priority of each fault data; among them, for the analysis of the probability distribution of the corresponding fault types for the fault operation data, the calculation is ; where L(n) is the fault type distribution function value of the fault operation data corresponding to the type number n; n(k), n x (k) and n(k) j correspond to the data values at each time point k in the current period of the fault operation data with the number n, the data values at each time point k in the next period of the fault operation data with the number n, and the average data values at each time point k in the historical period of the fault operation data with the number n respectively; by querying the fault type corresponding to the interval where the fault type distribution function value of the corresponding fault operation data is located, determine the fault type corresponding to the fault data existing in the faulty power equipment in the current period; by determining the historical influence weight of the corresponding fault type and the fault cause, match each fault cause corresponding to the fault type of the current fault operation data; according to the fault type of the corresponding fault operation data and each matching fault cause data, perform corresponding node connections in the device diagnosis knowledge graph respectively to obtain the inference path existing in the current fault operation data; based on the fault operation data inference path existing in the current faulty power equipment, analyze the proportion of each inference path of each fault operation data respectively, and the analysis calculation is ; where Y(g q (n)) corresponds to the fault proportion of the inference path number g q of the fault operation data with the number n; Fd(n) is the fault index of the fault operation data corresponding to the number n; L(n) i is the distribution function value of the i-th fault type corresponding to the fault operation data with the number n; W q,iIt is the weight value of the historical fault cause data corresponding to the fault type with the corresponding number i and the corresponding number q; since there may be multiple fault causes for the fault type corresponding to the single fault data, there will be multiple inference paths corresponding to a single fault data when constructing the inference path of the fault operation data. Sort the fault proportions of the inference paths corresponding to the fault operation data existing in the current faulty power equipment, construct the fault data table of the current faulty power equipment, sort it in descending order according to the fault proportion value of the inference path, and output the fault data table. Warn the corresponding fault operation data, corresponding fault type and fault cause of the faulty power equipment in the fault data table.

[0026] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic characteristics of the present invention. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be included in the present invention. Any reference signs in the claims should not be regarded as limiting the claims involved.

Claims

1. A power equipment fault location method based on a dynamic knowledge graph, characterized in that: The method includes the following steps: Determine the monitoring target of the power equipment, and collect the operation data of the target power equipment; Construct an adversarial learning framework for large-parameter models and small models, and perform output data processing on the data of large-parameter models and small models; Through the small model, perform feature analysis on the data of the target power equipment, combine the large-parameter model to perform parameter prediction on the data of the target power equipment, and determine the fault analysis data of the target power equipment; Based on the fault analysis data of the target power equipment by the large-parameter model and the small model, construct an equipment diagnosis knowledge graph, perform node analysis on the fault analysis data of the target equipment, determine the inference path of the fault data of the target power equipment, and output the fault data of the target power equipment.

2. The method for fault location of power equipment based on a dynamic knowledge graph according to claim 1, wherein: Obtain the operation node network of grid equipment through the grid control port, and determine the number and location data of the target power equipment; Retrieve the operation data packet of the target power equipment through the grid control port, and construct a periodic operation data table of the target power equipment.

3. The method for fault location of power equipment based on a dynamic knowledge graph according to claim 2, wherein: In the construction of the adversarial learning framework for large-parameter models and small models, plan the data output format through the types of each target power equipment and the corresponding operation data analysis scenarios, and formulate data conversion rules for the output data of the large-parameter model and the small models for data analysis of each corresponding target power equipment, so as to realize the alignment processing of the output data of the large-parameter model and the small models for data analysis of each corresponding target power equipment; Based on the output data of the large-parameter model and the small models for data analysis of each corresponding target power equipment after alignment processing, make a comparison and judgment on the output data manually, determine the difference situation of the output data of the large-parameter model and the small models for data analysis of each corresponding target power equipment, and based on the difference situation, adjust the model network data training structure or parameters of the large-parameter model, and update the adversarial learning data of the large-parameter model.

4. The method for fault location of power equipment based on a dynamic knowledge graph according to claim 3, wherein: Retrieve the periodic operation data table of each target power equipment, obtain various types of data of the periodic operation of the corresponding target power equipment within the corresponding period, and input the obtained data into the small models for data analysis of the corresponding target power equipment for feature processing; Construct a feature vector of the corresponding type of periodic operation data for the operation data of each time point in the corresponding type of period of the target power equipment. By integrating the eigenvectors corresponding to various types of operation data within a period of the target power equipment, a unified eigenvector space is constructed with the target power equipment as the central node to determine the set of eigenvectors of the periodic operation data of each target power equipment; the large-parameter model is used to predict various types of operation data of each target power equipment in the adjacent next period, and a set of eigenvectors of the predicted operation data of each target power equipment in the adjacent next period is constructed; combining the set of eigenvectors of the periodic operation data of each target power equipment in the current period and the set of eigenvectors of the predicted operation data of each target power equipment in the adjacent next period, a fault analysis is carried out for each target power equipment, and the faulty power equipment is determined based on the analysis results.

5. The method for fault location of power equipment based on a dynamic knowledge graph according to claim 4, characterized in that: Based on the large-parameter model combined with the small model for analyzing various types of operation data of the corresponding target power equipment, the faulty power equipment is determined according to the fault analysis results of the periodic operation data of each target power equipment; By constructing an equipment diagnosis knowledge graph, with the faulty equipment as the main node, various types of operation data within the period of the faulty equipment as the sub-nodes of layer A, the historical fault type data of the faulty equipment as the sub-nodes of layer B, and the historical fault cause data corresponding to the historical fault types of the faulty equipment as the sub-nodes of layer C; By analyzing the fault indices of various types of operation data within the period of the faulty power equipment, the periodic fault operation data of the faulty power equipment is determined; Based on the determined fault operation data, the corresponding fault operation data nodes are connected to the equipment main node in the equipment diagnosis knowledge graph to construct a fault data chain of the faulty power equipment; By performing a probability distribution analysis of the corresponding fault types on the determined fault operation data of the faulty power equipment, the fault type data of the current fault operation data is determined based on the analysis results; According to the historical fault cause data corresponding to the fault type data of the corresponding fault operation fault data for matching, and based on the matching results, a fault operation data inference path of the current faulty power equipment corresponding to the fault operation data is constructed; The fault proportions of the fault data inference paths corresponding to each fault data are analyzed respectively, and based on the analysis results, a fault data table of the current faulty power equipment is constructed respectively, and the priorities of each fault data are output.

6. A power equipment fault location system based on a dynamic knowledge graph, characterized in that: The system includes a power equipment monitoring module, a model construction and learning module, a fault determination module, and a fault analysis module; The power equipment monitoring module determines the monitoring target of the power equipment and collects the operation data of the target power equipment; the model construction and learning module constructs an adversarial learning framework of the large-parameter model and the small model, and performs output data processing on the data of the large-parameter model and the small model; The fault determination module analyzes the characteristics of the data of the target power equipment through the small model, combines the large-parameter model to predict the parameters of the data of the target power equipment, and determines the fault analysis data of the target power equipment; The fault analysis module constructs a device diagnosis knowledge graph based on the large-parameter model and the small model for the fault analysis data of the target power device, performs node analysis on the fault analysis data of the target device, determines the inference path of the fault data of the target power device, and outputs the fault data of the target power device.

7. The power equipment fault location system based on a dynamic knowledge graph according to claim 6, wherein: The power device monitoring module includes a target device determination unit and a data acquisition unit; The target device determination unit obtains the operation node network of the grid device according to the grid control port, and determines the number and positioning data of the target power device; The data acquisition unit retrieves the operation data packet of the target power device through the grid control port, and constructs a periodic target power device operation data table.

8. The power equipment fault location system based on a dynamic knowledge graph according to claim 7, characterized in that: The model construction and learning module includes a model data processing unit and a model parameter adjustment unit; The model data processing unit constructs a large-parameter model and a small model adversarial learning framework, plans the data output format through the analysis of the types of each target power device and the corresponding operation data analysis scenarios, and formulates data conversion rules for the output data of the large-parameter model and the small models for the data analysis of each corresponding target power device, so as to realize the alignment processing of the output data of the large-parameter model and the small models for the data analysis of each corresponding target power device; The model parameter adjustment unit, based on the output data of the large-parameter model and the small models for the data analysis of each corresponding target power device after alignment processing, makes a comparison and judgment on the output data manually, determines the difference situation between the output data of the large-parameter model and the small models for the data analysis of each corresponding target power device, and adjusts the model network data training structure or parameters of the large-parameter model based on the difference situation, and updates the adversarial learning data of the large-parameter model.

9. The power equipment fault location system based on a dynamic knowledge graph according to claim 8, wherein: The fault determination module includes an operation data feature processing unit and a faulty power device determination unit; The operation data feature processing unit retrieves the periodic target power device operation data table of each target power device, obtains various types of data of the corresponding target power device during the corresponding period, and inputs the obtained data into the small models for the data analysis of the corresponding target power device for feature processing; Constructs a feature vector of the corresponding type of periodic operation data for the operation data of each time point in the corresponding period of the target power device; The faulty power device determination unit integrates the feature vectors corresponding to various types of operation data within the period of the corresponding target power device, constructs a unified feature vector space with the corresponding target power device as the central node, and determines the set of feature vectors of the periodic operation data of each target power device; predicts various types of operation data within the next adjacent period of each target power device through the large-parameter model, and constructs a set of feature vectors of the predicted operation data within the next adjacent period of each target power device; combines the set of feature vectors of the periodic operation data of each target power device within the current period and the set of feature vectors of the predicted operation data within the next adjacent period, performs fault analysis on each corresponding target power device, and determines the faulty power device based on the analysis result.

10. The power equipment fault location system based on a dynamic knowledge graph according to claim 9, wherein: The fault analysis module includes a knowledge graph construction unit and a fault data feedback unit; The knowledge graph construction unit determines the faulty power equipment based on the fault analysis results of the periodic operation data of each target power equipment by combining a large parameter model with small models for analyzing various types of operation data of the corresponding target power equipment. By constructing an equipment diagnosis knowledge graph, with the faulty equipment as the main node, various types of operation data within the period of the faulty equipment as the sub-nodes of layer A, the historical fault type data of the faulty equipment as the sub-nodes of layer B, and the historical fault cause data corresponding to the historical fault types of the faulty equipment as the sub-nodes of layer C; By analyzing the fault indices of various types of operation data within the period of the faulty power equipment, the periodic fault operation data of the faulty power equipment is determined. The fault data feedback unit connects the corresponding fault operation data nodes with the equipment main node in the equipment diagnosis knowledge graph based on the determined fault operation data to construct a fault data chain for the faulty power equipment. By performing a probability distribution analysis of the corresponding fault types on the determined fault operation data of the faulty power equipment, the fault type data of the current fault operation data is determined based on the analysis results. Match according to the historical fault cause data corresponding to the fault type data of the corresponding fault operation fault data, and construct a fault operation data inference path for the corresponding fault operation data of the current faulty power equipment based on the matching results. Analyze the fault proportions of the fault data inference paths corresponding to each fault data respectively, and construct a fault data table for the current faulty power equipment based on the analysis results, and output the priorities of each fault data.

Images