Electric power material equipment defect analysis method and related device

Abstract

The invention belongs to the field of electric power material quality control, and discloses an electric power material equipment defect analysis method and a related device, and the method comprises the steps: obtaining knowledge graph quality detection sub-graphs of electric power material equipment of the same type, carrying out the clustering according to a spectral clustering algorithm, and determining defective electric power material equipment; obtaining a knowledge graph quality detection sub-graph set of the defective electric power material equipment, and performing frequent sub-graph analysis to obtain a maximum frequent sub-graph set of the defective electric power material equipment; and obtaining familial defects of the electric power material equipment according to the maximum frequent sub-graph set of the defective electric power material equipment. The defect identification of the electric power material equipment is realized, and the familial defects of the electric power material equipment are deeply analyzed, so that the state of the electric power material equipment can be more comprehensively known, potential risks can be timely found and processed, the operation and maintenance effect of the electric power material equipment is improved, and the service life of the electric power material equipment is prolonged.

Description

Technical Field

[0001] The present invention belongs to the field of power material quality control, and relates to a method for analyzing defects of power material equipment and related devices. Background Art

[0002] With the rapid advancement of power construction, the length of transmission lines of 35 kV and above has exceeded the milestone of millions of kilometers, the capacity of substation equipment has climbed to a new height of billions of kVA, and the annual grid investment amount has reached hundreds of billions of yuan. This large-scale power material equipment, like the "blood vessels" and "heart" of the power grid, once fails, may have a serious impact on the safe and stable operation of the power grid. Therefore, the defect detection work of power equipment is particularly important. It is an indispensable part of the power material quality control work and is directly related to the safety and stable operation of the power grid.

[0003] At present, various advanced sensing devices and data analysis algorithms have been widely and deeply applied to the online monitoring of electrical equipment, providing unprecedented technical support for the operation and maintenance of equipment. The effective application of these technical means not only greatly improves the operation and maintenance efficiency of equipment, enabling operation and maintenance personnel to quickly and accurately capture abnormal signals of equipment, but also can timely discover and handle equipment defects, eliminate potential safety hazards in the bud, and thus effectively avoid power grid accidents caused by equipment failures.

[0004] However, the current equipment detection work still has certain limitations, mainly focusing on the detection of single equipment or single faults, lacking comprehensive and in-depth analysis of equipment status. The existence of this limitation may lead to the neglect or misjudgment of potential risks of equipment, thereby affecting the operation and maintenance effect and service life of equipment. Summary of the Invention

[0005] The purpose of the present invention is to overcome the above-mentioned shortcomings of the prior art and provide a method for analyzing defects of power material equipment and related devices.

[0006] To achieve the above purpose, the present invention adopts the following technical solutions:

[0007] In the first aspect of the present invention, a method for analyzing defects of power material equipment is provided, including: obtaining knowledge graph quality detection subgraphs of each power material equipment of the same type, and clustering according to the spectral clustering algorithm to determine defective power material equipment; obtaining a set of knowledge graph quality detection subgraphs of the defective power material equipment, and performing frequent subgraph analysis to obtain a maximum frequent subgraph set of the defective power material equipment; obtaining the familial defects of the power material equipment according to the maximum frequent subgraph set of the defective power material equipment.

[0008] Optionally, the clustering according to the spectral clustering algorithm to determine defective power material equipment includes: obtaining quality inspection data of power material equipment based on the quality inspection sub-graph of the knowledge graph of power material equipment, and using it as clustering sample points; using the spectral clustering algorithm based on the symmetric Laplacian matrix to cluster the clustering sample points of each power material equipment of the same type to obtain several clustering clusters; analyzing the equipment performance of several clustering clusters, and obtaining the power material equipment corresponding to the clustering sample points within the clustering cluster with defective equipment performance, so as to obtain defective power material equipment.

[0009] Optionally, the performing frequent sub-graph analysis to obtain the maximum frequent sub-graph set of defective power material equipment includes: obtaining several candidate sub-graphs according to the quality inspection sub-graph set of the knowledge graph of defective power material equipment by using the fast frequent sub-graph mining method; using the pruning strategy based on support degree combined with a preset minimum support degree threshold to prune several candidate sub-graphs to obtain several initially screened candidate sub-graphs; using the sub-graph pruning and inductive network - maximum frequent sub-graph mining method to perform pruning and expansion processing on several initially screened candidate sub-graphs to obtain the maximum frequent sub-graph set of defective power material equipment.

[0010] Optionally, it further includes: obtaining the quality inspection sub-graph of the union of all power material equipment within the preset area of the defective power material equipment as the associated fault analysis sub-graph; performing dense sub-graph mining on the associated fault analysis sub-graph to obtain several dense sub-graphs; setting each power material equipment included in the dense sub-graph containing the defective power material equipment among several dense sub-graphs as the first-level controlled power material equipment.

[0011] Optionally, it further includes: obtaining the power material equipment with the shortest distance to the defective power material equipment within the dense sub-graph containing the defective power material equipment among several dense sub-graphs, and setting it as the second-level controlled power material equipment.

[0012] Optionally, the performing dense sub-graph mining on the associated fault analysis sub-graph includes: using the edge density as the density of the graph, and performing top-k dense sub-graph mining on the associated fault analysis sub-graph by the branch and bound method.

[0013] In the second aspect of the present invention, a power material equipment defect analysis system is provided, including: a defective equipment identification module, configured to obtain the quality inspection sub-graph of each power material equipment of the same type, and determine defective power material equipment according to the spectral clustering algorithm; a frequent sub-graph analysis module, configured to obtain the quality inspection sub-graph set of the knowledge graph of defective power material equipment, and perform frequent sub-graph analysis to obtain the maximum frequent sub-graph set of defective power material equipment; a familial defect analysis module, configured to obtain the familial defects of power material equipment according to the maximum frequent sub-graph set of defective power material equipment.

[0014] Optionally, it also includes: an associated subgraph acquisition module, used to obtain a knowledge graph quality detection subgraph that is combined with all power material equipment in a preset area of ​​defective power material equipment as an associated fault analysis subgraph; a dense subgraph analysis module, used to perform dense subgraph mining on the associated fault analysis subgraph to obtain several dense subgraphs; and a first-level control setting module, used to set each power material equipment contained in a dense subgraph containing defective power material equipment among several dense subgraphs as a first-level controlled power material equipment.

[0015] According to a third aspect of the present invention, a computer device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned method for analyzing defects in electric power materials and equipment when executing the computer program.

[0016] According to a fourth aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned method for analyzing defects in electric power materials and equipment are implemented.

[0017] Compared with the prior art, the present invention has the following beneficial effects:

[0018] The method for analyzing defects in electric power materials and equipment of the present invention first obtains the knowledge graph quality detection subgraphs of each electric power material and equipment of the same type, and uses the spectral clustering algorithm for clustering. It can cluster electric power material and equipment with similar characteristics based on the structural information of the graph, thereby accurately identifying defective electric power material and equipment. Then, the knowledge graph quality detection subgraph set of defective electric power material and equipment is obtained, and frequent subgraph analysis is performed. This step can mine the subgraph patterns that exist in common in defective electric power material and equipment, that is, the maximum frequent subgraph set. These subgraph patterns reflect the familial characteristics of the defects of defective electric power material and equipment, and are important indicators of potential risks of the equipment. Through this method, not only the defect identification of electric power material and equipment is realized, but also the familial defects of electric power material and equipment are deeply analyzed, which helps to have a more comprehensive understanding of the status of electric power material and equipment, timely discover and deal with potential risks, thereby improving the operation and maintenance effect and service life of electric power material and equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 The present invention is a flowchart of a method for analyzing defects in electric power materials and equipment according to an embodiment of the present invention.

[0020] Figure 2 This is a structural block diagram of a power material equipment defect analysis system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0021] To enable those skilled in the art to better understand the solution of the present invention, the following will clearly and completely describe the technical solution in the embodiments of the present invention in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative work shall fall within the protection scope of the present invention.

[0022] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products, or devices.

[0023] The following further describes the present invention in detail with reference to the accompanying drawings:

[0024] See Figure 1 , in an embodiment of the present invention, a method for analyzing defects of power material equipment is provided, which can not only effectively identify defective power material equipment in power material equipment, but also discover the familial defects of power material equipment, and further discover the potential risks of power material equipment, providing a basis for operation and maintenance.

[0025] Specifically, the method for analyzing defects of power material equipment in the present invention includes the following steps:

[0026] S1: Obtain the knowledge graph quality detection subgraphs of each power material equipment of the same type, and perform clustering according to the spectral clustering algorithm to determine the defective power material equipment.

[0027] S2: Obtain the set of knowledge graph quality detection subgraphs of defective power material equipment, and perform frequent subgraph analysis to obtain the maximum frequent subgraph set of defective power material equipment.

[0028] S3: Obtain the familial defects of power material equipment according to the maximum frequent subgraph set of defective power material equipment.

[0029] The defect analysis method for power material equipment of the present invention first obtains the knowledge graph quality detection subgraphs of various power material equipment of the same type and uses the spectral clustering algorithm for clustering. Based on the structural information of the graph, it can cluster power material equipment with similar characteristics together, thus accurately identifying defective power material equipment. Then, it obtains the set of knowledge graph quality detection subgraphs of defective power material equipment and conducts frequent subgraph analysis. This step can excavate the common subgraph patterns existing in defective power material equipment, that is, the maximum frequent subgraph set. These subgraph patterns reflect the familial characteristics of the defects of defective power material equipment and are important indicators of the potential risks of the equipment. Through this method, not only the defect identification of power material equipment is realized, but also the familial defects of power material equipment are deeply analyzed, which helps to more comprehensively understand the state of power material equipment, timely discover and handle potential risks, and thus improve the operation and maintenance effect and service life of power material equipment.

[0030] Explanatory, the basic data of the knowledge graph quality detection subgraph of power material equipment includes the quality detection data generated during the production and operation of power material equipment, as well as basic information such as the manufacturer, production batch, and operation location of power material equipment. These basic data are managed in the form of a knowledge graph. On this basis, according to business requirements, the knowledge graph quality detection subgraphs of relevant power material equipment can be extracted. For example, to conduct defect analysis on a transformer, the knowledge graph quality detection subgraph of the transformer is screened out.

[0031] Explanatory, various power material equipment of the same type generally refers to power material equipment with the same equipment type and model, or it can also be power material equipment with only the same equipment type.

[0032] In a possible implementation manner, the clustering according to the spectral clustering algorithm to determine defective power material equipment includes: obtaining the quality detection data of power material equipment based on the knowledge graph quality detection subgraph of power material equipment and using it as clustering sample points; using the spectral clustering algorithm based on the symmetric Laplacian matrix to cluster the clustering sample points of various power material equipment of the same type to obtain several clustering clusters; conducting equipment performance analysis on the several clustering clusters and obtaining the power material equipment corresponding to the clustering sample points within the clustering cluster with defective equipment performance, so as to obtain defective power material equipment.

[0033] Explanatory, the spectral clustering algorithm is a clustering method based on graph theory. It regards data points as vertices in a graph and uses the similarity between vertices as the weight of edges to construct a weighted graph. Then, the spectral clustering algorithm divides the graph into several optimal subgraphs by cutting the graph, so that the similarity of vertices within the subgraph is as high as possible, and the similarity between subgraphs is as low as possible, thus achieving the purpose of clustering.

[0034] In this embodiment, the spectral clustering algorithm adopted is the spectral clustering algorithm based on the symmetric Laplacian matrix. The spectral clustering algorithm based on the symmetric Laplacian matrix is a clustering method based on graph theory. Its core idea is to perform eigen-decomposition using the Laplacian matrix of the graph, thereby realizing data clustering.

[0035] Exemplarily, the steps for clustering the knowledge graph quality detection subgraphs of each power material equipment include:

[0036] Step 1: According to the knowledge graph quality detection subgraphs of power material equipment, obtain the quality detection data of power material equipment. Exemplarily, taking a transformer as an example, based on the knowledge graph quality detection subgraph of the transformer, the quality detection data of each transformer can be obtained. For example: for Transformer 1, the infrared temperature measurement result is 50°C, and the partial discharge detection is 50 pC; for Transformer 2, the infrared temperature measurement result is 60°C, and the partial discharge detection is 70 pC. Finally, generate clustering sample points: X1 = {50, 50} which is the clustering sample point representing Transformer 1 and X2 = {60, 70} which is the clustering sample point representing Transformer 2. Finally, generate n clustering sample points X = {x1, x2, …, x n}, and preset the number of clusters k according to historical experience or expert experience. Among them, the number of clusters k can be adjusted according to business needs, and the initial value can be taken as 2 or 3. Then use the Gaussian kernel function to calculate the similarity matrix W:

[0037]

[0038] where W ij is the similarity between the i-th clustering sample point x i and the j-th clustering sample point x j , W ji is the similarity between the j-th clustering sample point x j and the i-th clustering sample point x i , and σ is a preset hyperparameter.

[0039] Step 2: Calculate the degree matrix D, and calculate the Laplacian matrix L = D -1 / 2 (D - W)D -1 / 2 .

[0040] Step 3: Calculate the eigenvalues of the Laplacian matrix L, sort the eigenvalues from smallest to largest, then take the first k eigenvalues, and calculate the eigenvectors corresponding to the first k eigenvalues. Combine the eigenvectors corresponding to the above k eigenvalues to form an eigenmatrix U = {u1, u2, …, u k} where U ∈ R n×k .

[0041] Step 4: Let the new sample point y i be the i-th row vector of the eigenmatrix U; for i = 1, 2, …, n, take yi Normalize them successively so that |y i | = 1; then use the k-means algorithm (K-Means algorithm) to cluster the new sample points Y = {y1, y2, …, y n} into a cluster set Y = {C1, C2, …, C k}.

[0042] Step Five: Output clustering clusters: A1, A2, …, A k ; where A i = {j|y j ∈ C i}, representing the i-th clustering cluster. The performance of the power material equipment corresponding to the clustering sample points in the same clustering cluster is similar. Based on this, perform equipment performance analysis on several clustering clusters, and preferentially select the clustering clusters with fewer clustering sample points for analysis because under normal circumstances, most power material equipment is in a normal state. Then, obtain the power material equipment corresponding to the clustering sample points in the clustering cluster with defective equipment performance, and obtain defective power material equipment.

[0043] In a possible implementation manner, the performing frequent subgraph analysis to obtain the maximum frequent subgraph set of defective power material equipment includes: according to the knowledge graph quality detection subgraph set of defective power material equipment, using a fast frequent subgraph mining method to obtain several candidate subgraphs; using a pruning strategy based on support degree combined with a preset minimum support threshold to prune several candidate subgraphs to obtain several initially screened candidate subgraphs; using a subgraph pruning and inductive network - maximum frequent subgraph mining method to perform pruning and expansion processing on several initially screened candidate subgraphs to obtain the maximum frequent subgraph set of defective power material equipment.

[0044] Explanatorily, according to the knowledge graph quality detection subgraph set of defective power material equipment, obtain the familial defect information of power material equipment through frequent subgraph analysis.

[0045] Frequent subgraph mining is a method of discovering a set of common substructures in a set of graphs. During the frequent subgraph mining process, the support degree is defined as: given a set family of graphs, the support degree s(g) of subgraph g is defined as the percentage of all graphs that contain it. The goal of frequent subgraph mining is to find subgraph g such that all s(g) ≥ minsup, where minsup is the preset minimum support threshold.

[0046] The fast frequent subgraph mining method (Fast Frequent Subgraph Mining, FFSM) adopts a vertical search mode to significantly improve the efficiency of frequent subgraph mining by solving potential subgraph isomorphism problems and reducing the generation of redundant candidate subgraphs, and is applicable to various application scenarios that require mining frequent subgraphs.

[0047] First, the FFSM method is used to obtain several candidate subgraphs from the knowledge graph quality detection subgraph set of defective power material equipment. Specifically, the FFSM method first obtains a canonical encoding using a canonical adjacency matrix (CAM), and then uses the FFSM_Join function and the FFSM_Extension function to generate candidate subgraphs. Secondly, pruning is performed on the generated candidate subgraphs. In this embodiment, a pruning strategy based on support is combined with a preset minimum support threshold for pruning to filter out non-frequent subgraphs and subgraphs whose canonical adjacency matrices are not sub-optimal, ultimately achieving the purpose of reducing the number of candidate subgraphs and obtaining several initially screened candidate subgraphs.

[0048] Next, the Subgraph Pruning and Induction Network - Maximal Frequent subGraph Mining (SPIN - MGM) method is used to perform pruning and expansion processing on several initially screened candidate subgraphs to obtain the maximal frequent subgraph set of defective power material equipment. Among them, the SPIN - MGM method is implemented in two stages: the SPIN stage: pruning and expanding the initially screened candidate subgraphs; the MGM stage: mining the maximal frequent subgraphs that meet the support threshold in the optimized initially screened candidate subgraphs, that is, the frequent subgraphs not contained by other frequent subgraphs.

[0049] Explanatorily, based on the maximal frequent subgraph set of defective power material equipment, the similar structure of the knowledge graph quality detection subgraph of defective power material equipment can be obtained. By viewing each maximal frequent subgraph based on this similar structure, the family defect information of defective power material equipment can be obtained. Exemplarily, the obtained family defect information can be: a certain batch of transformers produced by a certain manufacturer has family defects, or the transformers produced by a certain manufacturer have family defects in infrared temperature measurement, etc.

[0050] In a possible implementation manner, the power material equipment defect analysis method further includes: obtaining the knowledge graph quality detection subgraph of all power material equipment combined within a preset area of the defective power material equipment as an associated fault analysis subgraph; performing dense subgraph mining on the associated fault analysis subgraph to obtain several dense subgraphs; setting each power material equipment included in the dense subgraph containing the defective power material equipment among the several dense subgraphs as a first-level controlled power material equipment.

[0051] Explanatory, the occurrence of power material equipment failures is often interrelated. The emergence of multi-source uncertain failures often leads to abnormal monitoring values of sensors of nearby equipment. Therefore, a dense subgraph mining algorithm is used to discover concurrent failures from a large amount of graph data. Specifically, according to the defective power material equipment extracted by the spectral clustering algorithm, a knowledge graph quality inspection subgraph of all power material equipment within the preset area of the defective power material equipment is extracted from the knowledge graph for the analysis and early warning of associated failures, and then the power material equipment prone to associated failures is determined through dense subgraph analysis. A dense subgraph refers to a sub-region with relatively dense internal edges in a graph, which is generally regarded as the core part of the graph structure and contains relatively concentrated knowledge. On this basis, it is determined that the power material equipment in the dense subgraph is the key equipment prone to associated failures, and the monitoring intensity of these power material equipment needs to be strengthened to prevent more associated failures from occurring. Exemplarily, the preset area is generally the same substation, the same transmission line or the same distribution transformer area, etc.

[0052] In a possible implementation manner, the dense subgraph mining of the associated failure analysis subgraph includes: using the edge density as the density of the graph, and performing top-k dense subgraph mining on the associated failure analysis subgraph by the branch and bound method.

[0053] Explanatory, the edge density d(S) is used to measure the density of the graph S:

[0054]

[0055] where |E(S)| represents the number of edges of the graph S, and |V(S)| represents the number of nodes of the graph S.

[0056] The specific steps for performing top-k dense subgraph mining on the associated failure analysis subgraph by the branch and bound method are as follows:

[0057] Step (1): Initialize the priority queue. Create a priority queue Q for storing the nodes to be searched, and sort them according to the estimated value of the density index of the subgraph represented by the nodes. The higher the estimated value, the higher the priority.

[0058] Step (2): Create the root node. Create the root node r, representing the entire graph G, calculate the estimated value f(r) of the density index of the root node, and add the root node r to the priority queue Q. At the same time, initialize a set R for storing the top-k dense subgraphs that have been found. Initially, the set R is empty.

[0059] Step (3): Initialize the upper bound. According to the characteristics of the graph G, calculate an initial global upper bound U, and use the edge density of the entire graph G as the initial global upper bound U.

[0060] Step (4): Branch node selection. Take out the node n with the highest priority from the priority queue Q.

[0061] Step (5): Branching strategy. Branch the node n to generate its child nodes. The main branching strategies are as follows: 1. Add node branch: Select a node v in the graph G that is not in the subgraph represented by the node n, and add the node v to the subgraph to generate a new child node. 2. Delete node branch: Select a node u in the subgraph represented by the node n, and delete the node u from the subgraph to generate a new child node. 3. Add edge branch: Select an edge e in the graph G that connects two nodes in the subgraph represented by the node n, and add the edge e to the subgraph to generate a new child node. 4. Delete edge branch: Select an edge f in the subgraph represented by the node n, and delete the edge f from the subgraph to generate a new child node.

[0062] Step (6): Bound calculation. For each child node c, calculate the actual density index d(c) of its corresponding subgraph as the lower bound. Estimate an upper bound U(c) according to the structure of the child node c and the remaining part of the graph G. When adding a node branch, the maximum edge density of the subgraph after adding all possible edges can be estimated as the upper bound based on the connection situation between the remaining nodes and the current subgraph.

[0063] Step (7): Pruning. Compare the lower bound d(c) of the child node c with the density index value of the k-th densest subgraph known currently (i.e., the minimum density index value in the set R). If d(c) is less than this value, prune the child node c and do not continue the search. Compare the upper bound U(c) of the child node c with the global upper bound U. If the upper bound U(c) of the child node c is not greater than the global upper bound U, prune the child node c.

[0064] Step (8): Solution checking and updating. Check whether the subgraph represented by the child node c satisfies the constraint conditions such as the node number limit and the edge weight limit. If the subgraph represented by the child node c is feasible and its density index is greater than the minimum density index value in the set R, delete the subgraph corresponding to the minimum density index value in the set R, and add the subgraph represented by the child node c to the set R. At the same time, update the global upper bound U to the minimum density index value in the set R.

[0065] Step (9): Repeat steps (4) to (8) until the priority queue Q is empty. At this time, the top-k densest subgraphs are stored in the set R.

[0066] Explanatorily, the top-k dense subgraphs represent the k subgraphs with the highest density, indicating that the power material equipment within them is closely connected and more prone to concurrent failures. Therefore, each power material equipment included in the dense subgraph containing defective power material equipment is set as a first-level controlled power material equipment and managed as a key control area. Temporary monitoring measures need to be added to promptly detect and handle concurrent failures.

[0067] In a possible implementation manner, the power material equipment defect analysis method further includes: obtaining the power material equipment with the shortest distance to the defective power material equipment within the dense subgraph containing the defective power material equipment among several dense subgraphs, and setting it as the second-level controlled power material equipment.

[0068] Explanatorily, calculate the shortest distance between the defective power material equipment and each power material equipment within the dense subgraph containing the defective power material equipment, excluding the distance between the defective power material equipment and itself. And consider the power material equipment corresponding to the shortest shortest distance as the power material equipment most likely to have associated failures, and set it as the second-level controlled power material equipment to strengthen monitoring to prevent failures. Among them, the monitoring level of the second-level controlled power material equipment should be higher than that of the first-level controlled power material equipment.

[0069] The power material equipment defect analysis method of the present invention first realizes the identification of defective power material equipment based on the spectral clustering algorithm. Then, on the basis of the knowledge graph quality detection subset of defective power material equipment, the identification of familial defects of defective power material equipment is realized through frequent subgraph analysis. Moreover, obtain the knowledge graph quality detection subgraph of the union of all power material equipment within the preset area of the defective power material equipment as the associated failure analysis subgraph, and on the basis of the associated failure analysis subgraph, the identification of power material equipment with high-risk associated failures is realized through dense subgraph analysis and shortest path calculation. It fully excavates the information of the power material equipment knowledge graph and effectively improves the quality control level of power material equipment.

[0070] The following is an apparatus embodiment of the present invention, which can be used to execute the method embodiment of the present invention. For details not disclosed in the apparatus embodiment, please refer to the method embodiment of the present invention.

[0071] See Figure 2 , in another embodiment of the present invention, a power material equipment defect analysis system is provided, which can be used to implement the above-mentioned power material equipment defect analysis method. Specifically, the power material equipment defect analysis system includes a defective equipment identification module, a frequent subgraph analysis module, and a familial defect analysis module.

[0072] Among them, the defective equipment identification module is used to obtain the knowledge graph quality inspection subgraphs of various power material equipment of the same type, perform clustering according to the spectral clustering algorithm, and determine the defective power material equipment; the frequent subgraph analysis module is used to obtain the set of knowledge graph quality inspection subgraphs of the defective power material equipment, and perform frequent subgraph analysis to obtain the maximum frequent subgraph set of the defective power material equipment; the familial defect analysis module is used to obtain the familial defects of the power material equipment according to the maximum frequent subgraph set of the defective power material equipment.

[0073] In a possible implementation manner, the clustering according to the spectral clustering algorithm to determine the defective power material equipment includes: obtaining the quality inspection data of the power material equipment according to the knowledge graph quality inspection subgraph of the power material equipment, and using it as the clustering sample points; using the spectral clustering algorithm based on the symmetric Laplacian matrix to cluster the clustering sample points of various power material equipment of the same type to obtain several clustering clusters; performing equipment performance analysis on the several clustering clusters, and obtaining the power material equipment corresponding to the clustering sample points within the clustering cluster with defective equipment performance, so as to obtain the defective power material equipment.

[0074] In a possible implementation manner, the performing frequent subgraph analysis to obtain the maximum frequent subgraph set of the defective power material equipment includes: obtaining several candidate subgraphs according to the set of knowledge graph quality inspection subgraphs of the defective power material equipment by using the fast frequent subgraph mining method; using the pruning strategy based on support degree combined with the preset minimum support degree threshold to prune the several candidate subgraphs to obtain several preliminarily screened candidate subgraphs; using the subgraph pruning and inductive network - maximum frequent subgraph mining method to perform pruning and expansion processing on the several preliminarily screened candidate subgraphs to obtain the maximum frequent subgraph set of the defective power material equipment.

[0075] In a possible implementation manner, the power material equipment defect analysis system further includes: an associated subgraph acquisition module, a dense subgraph analysis module, and a first - level control setting module.

[0076] Among them, the associated subgraph acquisition module is used to obtain the knowledge graph quality inspection subgraph of the union of all power material equipment within the preset area of the defective power material equipment as the associated fault analysis subgraph; the dense subgraph analysis module is used to perform dense subgraph mining on the associated fault analysis subgraph to obtain several dense subgraphs; the first - level control setting module is used to set each power material equipment included in the dense subgraph containing the defective power material equipment among the several dense subgraphs as the first - level controlled power material equipment.

[0077] In a possible implementation, the power material equipment defect analysis system further includes: a secondary control setting module. The secondary control setting module is used to obtain the power material equipment with the shortest distance to the defective power material equipment within the dense subgraph containing the defective power material equipment among several dense subgraphs, and set it as the secondary control power material equipment.

[0078] In a possible implementation, the dense subgraph mining of the associated fault analysis subgraph includes: using the edge density as the graph density, and performing top-k dense subgraph mining on the associated fault analysis subgraph through the branch and bound method.

[0079] All relevant contents of each step involved in the embodiments of the foregoing power material equipment defect analysis method can be cited in the function descriptions of the corresponding functional modules of the power material equipment defect analysis system in the embodiments of the present invention, and will not be elaborated here again.

[0080] The division of modules in the embodiments of the present invention is illustrative, only a logical function division. In actual implementation, there may be other division methods. In addition, in each embodiment of the present invention, the functional modules can be integrated in a processor, or can exist independently physically, or two or more modules can be integrated in one module. The above integrated modules can be implemented in the form of hardware or in the form of software functional modules.

[0081] In another embodiment of the present invention, a computer device is provided. The computer device includes a processor and a memory. The memory is used to store a computer program, and the computer program includes program instructions. The processor is used to execute the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, and is suitable for implementing one or more instructions. Specifically, it is suitable for loading and executing one or more instructions in the computer storage medium to implement the corresponding method flow or corresponding function; the processor described in the embodiments of the present invention can be used for the operation of the power material equipment defect analysis method.

[0082] In another embodiment of the present invention, the present invention further provides a storage medium, specifically a computer-readable storage medium (Memory). The computer-readable storage medium is a memory device in a computer device and is used to store programs and data. It can be understood that the computer-readable storage medium here can include both the built-in storage medium in the computer device and, of course, the extended storage medium supported by the computer device. The computer-readable storage medium provides a storage space, and the operating system of the terminal is stored in this storage space. Moreover, one or more instructions suitable for being loaded and executed by the processor are stored in this storage space, and these instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory. One or more instructions stored in the computer-readable storage medium can be loaded and executed by the processor to implement the corresponding steps of the method for analyzing defects in power material equipment in the above embodiments.

[0083] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memories, CD-ROMs, optical memories, etc.) containing computer-usable program codes.

[0084] The present invention is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of the flows and / or blocks in the flowchart and / or block diagram can also be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0085] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device, and the instruction device implements the functions in Figure 1 one flow or multiple flows and / or blocksFigure 1 The functions specified in one or more boxes.

[0086] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process. Thus, the instructions executed on the computer or other programmable device provide for implementing the steps of the functions specified in one or more processes and / or boxes Figure 1 One process or more processes and / or boxes Figure 1 The steps of the functions specified in one or more boxes.

[0087] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: the specific implementation manners of the present invention can still be modified or equivalently replaced, and any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered by the protection scope of the claims of the present invention.

Claims

1. A method for analyzing defects of power material equipment, characterized in that, Including: Obtain the knowledge graph quality detection subgraphs of each power material equipment of the same type, and perform clustering according to the spectral clustering algorithm to determine the defective power material equipment; Obtain the set of knowledge graph quality detection subgraphs of the defective power material equipment, and perform frequent subgraph analysis to obtain the maximum frequent subgraph set of the defective power material equipment; Obtain the familial defects of the power material equipment according to the maximum frequent subgraph set of the defective power material equipment.

2. The power material equipment defect analysis method according to claim 1, characterized in that The clustering according to the spectral clustering algorithm to determine the defective power material equipment includes: According to the knowledge graph quality detection subgraph of the power material equipment, obtain the quality detection data of the power material equipment and use it as the clustering sample points; Adopt the spectral clustering algorithm based on the symmetric Laplacian matrix to cluster the clustering sample points of each power material equipment of the same type to obtain several clustering clusters; Perform equipment performance analysis on several clustering clusters, and obtain the power material equipment corresponding to the clustering sample points within the clustering cluster with defective equipment performance, so as to obtain the defective power material equipment.

3. The power material equipment defect analysis method according to claim 1, wherein The performing frequent subgraph analysis to obtain the maximum frequent subgraph set of the defective power material equipment includes: According to the set of knowledge graph quality detection subgraphs of the defective power material equipment, use the fast frequent subgraph mining method to obtain several candidate subgraphs; Adopt the pruning strategy based on support combined with the preset minimum support threshold to prune several candidate subgraphs to obtain several initially screened candidate subgraphs; Adopt the subgraph pruning and induction network - maximum frequent subgraph mining method to perform pruning and expansion processing on several initially screened candidate subgraphs to obtain the maximum frequent subgraph set of the defective power material equipment.

4. The power material equipment defect analysis method according to claim 1, characterized in that Also including: Obtain the knowledge graph quality detection subgraph jointly formed by all power material equipment within the preset area of the defective power material equipment as the associated fault analysis subgraph; Perform dense subgraph mining on the associated fault analysis subgraph to obtain several dense subgraphs; Set each power material equipment included in the dense subgraph containing the defective power material equipment among several dense subgraphs as the first-level controlled power material equipment.

5. The power material equipment defect analysis method according to claim 4, wherein Also including: Obtain the power material equipment with the shortest shortest distance between it and the defective power material equipment within the dense subgraph containing the defective power material equipment among several dense subgraphs, and set it as the second-level controlled power material equipment.

6. The power material equipment defect analysis method according to claim 5, wherein The performing dense subgraph mining on the associated fault analysis subgraph includes: using the edge density as the graph density, and performing top-k dense subgraph mining on the associated fault analysis subgraph through the branch and bound method.

7. A power material equipment defect analysis system, characterized in that, Including: Defective equipment identification module, used to obtain the knowledge graph quality detection subgraphs of each power material equipment of the same type, and perform clustering according to the spectral clustering algorithm to determine the defective power material equipment; Frequent subgraph analysis module, used to obtain the set of knowledge graph quality detection subgraphs of the defective power material equipment, and perform frequent subgraph analysis to obtain the maximum frequent subgraph set of the defective power material equipment; Familial defect analysis module, used to obtain the familial defects of the power material equipment according to the maximum frequent subgraph set of the defective power material equipment.

8. The power material equipment defect analysis system according to claim 7, characterized in that Also including: Associated subgraph acquisition module, used to obtain the knowledge graph quality detection subgraph jointly formed by all power material equipment within the preset area of the defective power material equipment as the associated fault analysis subgraph; A dense subgraph analysis module is used to mine dense subgraphs from the associated fault analysis subgraph to obtain several dense subgraphs; A first-level control setting module is used to set each power material equipment included in the dense subgraph containing defective power material equipment among several dense subgraphs as first-level controlled power material equipment.

9. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, the steps of the power material equipment defect analysis method according to any one of claims 1 to 6 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, the steps of the power material equipment defect analysis method according to any one of claims 1 to 6 are implemented.

Images