CN-121095008-B - Multi-source data fusion power knowledge processing method, device, medium and equipment

CN121095008BCN 121095008 BCN121095008 BCN 121095008BCN-121095008-B

Abstract

The application provides a method, a device, a medium and equipment for processing electric power knowledge of multi-source data fusion, which relate to the field of electric power data processing and comprise the steps of obtaining graph structure data corresponding to each preset data source of a target electric power system; the method comprises the steps of obtaining standard function diagram structure data corresponding to each preset data source by carrying out function mapping on each node in each diagram structure data, determining each target equipment node in different preset data sources according to each standard function diagram structure data and a preset diagram similarity algorithm, and fusing the diagram structure data corresponding to each preset data source by taking each target equipment node as a fusion node to obtain a multi-data-source fusion power knowledge diagram. The integrated multi-data source integrated power knowledge graph has the characteristics of consistent data, uniform semantics and complete topology.

Inventors

WANG XUXIAN
CHEN DANNI
CHEN LIANG
WANG GUORUI
PEI QIUGEN
JIANG JIANG
WU SHEN
ZENG LIANGBO
SU HUAQUAN
YE YANGWEI

Assignees

广东电网有限责任公司

Dates

Publication Date: 20260512
Application Date: 20250828

Claims (9)

1. A method for processing power knowledge of multi-source data fusion, the method comprising: s100, obtaining graph structure data corresponding to each preset data source of a target power system, wherein each graph structure data comprises at least one target equipment node, and node identifiers of the target equipment nodes contained in each graph structure data are different; s200, performing function mapping on each node in each graph structure data to obtain standard function graph structure data corresponding to each preset data source, wherein each node in the standard function graph structure data is a function node, and the function node is a node named according to the function; S300, determining each target equipment node in different preset data sources according to each standard functional diagram structure data and a preset diagram similarity algorithm, wherein the preset diagram similarity algorithm determines whether the preset equipment node is a target equipment node according to the similarity between network structures and the similarity between node attributes in different standard functional diagram structure data of the preset equipment node; S400, fusing the graph structure data corresponding to each preset data source by taking each target equipment node as a fusion node so as to obtain a multi-data source fusion power knowledge graph; Step S300 includes: S310, acquiring an initial similarity matrix, wherein the initial similarity matrix is n multiplied by n, n is the total number of nodes contained in all the image structure data, the similarity of key elements in the initial similarity matrix is 1, the similarity of non-key elements is 0, the similarity of key elements between the i-th row and the i-th column nodes is the similarity of non-key elements between the i-th row and the j-th column nodes, i=1, 2, n, j=1, 2, n, i is not equal to j; S320, determining the current similarity value of each non-key element according to the initial similarity matrix and a preset calculation rule to obtain an intermediate similarity matrix; S330, obtaining an average similarity difference value of the intermediate similarity matrix; S340, if the average similarity difference value is equal to or greater than a preset difference threshold value, determining the current similarity value of each non-key element according to the intermediate similarity matrix and a preset calculation rule to update the intermediate similarity matrix, and jumping to step S330 until the average similarity difference value is less than the preset difference threshold value, and determining the current intermediate similarity matrix as a target similarity matrix; S350, determining the target equipment node according to the similarity between the target similarity matrix and the node attribute.
2. The method for power knowledge processing with multi-source data fusion according to claim 1, wherein step S320 comprises: S321, obtaining the sum of the similarity between each node adjacent to the ith row node and each node adjacent to the jth column node corresponding to each non-key element, wherein the similarity between each node adjacent to the ith row node and each node adjacent to the jth column node is a similarity value of a corresponding position in an initial similarity matrix; S322, obtaining a current similarity value of each non-key element according to the sum of the similarity between each node adjacent to the ith row node and each node adjacent to the jth column node corresponding to each non-key element, a preset attenuation factor and the product of the number of the nodes adjacent to the ith row node and the number of the nodes adjacent to the jth column node of the non-key element, so as to obtain an intermediate similarity matrix.
3. The method for processing the power knowledge based on the multi-source data fusion according to claim 2, wherein the current similarity value of each non-critical element is proportional to the sum of the similarity between each node adjacent to the ith row node and each node adjacent to the jth column node corresponding to the non-critical element, is proportional to a preset attenuation factor, and is inversely proportional to the product of the number of nodes adjacent to the ith row node and the number of nodes adjacent to the jth column node of the non-critical element.
4. The method for power knowledge processing with multi-source data fusion according to claim 2, wherein step S350 comprises: s351, obtaining the similarity between any two preset equipment nodes in the target similarity matrix, wherein the preset equipment nodes are nodes with the same functions as the target equipment nodes, and the target equipment nodes are any one of the preset equipment nodes; S352, if the similarity between any two preset equipment nodes is greater than a first preset similarity threshold, determining that the corresponding two nodes are key equipment node pairs; S353, obtaining the similarity between the node attributes of each key device node pair; S354, obtaining key similarity between each key equipment node pair according to the similarity of each key equipment node pair in the target similarity matrix and the similarity between node attributes, wherein the key similarity is in direct proportion to the similarity of the key equipment node pair in the target similarity matrix; S355, if the key similarity between any key node pair is greater than a preset key similarity threshold, determining that two preset device nodes contained in the key node pair are the same target device node.
5. The method for power knowledge processing for multi-source data fusion according to claim 4, further comprising, after step S351: s356, if the similarity between any two preset device nodes is smaller than the first preset similarity threshold and the second preset similarity threshold, determining that the corresponding two nodes are intermediate device node pairs; s357, if the absolute value of the adjacent node quantity difference between the two preset equipment nodes contained in the intermediate equipment node pair is larger than the preset adjacent node quantity difference threshold, obtaining the similarity between the node attributes of the intermediate equipment node pair; S358, obtaining intermediate similarity between each intermediate equipment node pair according to the similarity of each intermediate equipment node pair in the target similarity matrix and the similarity between node attributes, wherein the intermediate similarity is obtained by weighted summation of the similarity of the intermediate equipment node in the target similarity matrix and the similarity between node attributes of the intermediate equipment node, and the weight of the similarity between node attributes of the intermediate equipment node is greater than the weight of the similarity of the intermediate equipment node in the target similarity matrix; S359, if the intermediate similarity between any intermediate node pair is greater than the preset intermediate similarity threshold, determining that two preset device nodes included in the intermediate node pair are the same target device node.
6. The method for processing the power knowledge based on multi-source data fusion according to claim 4 or 5, wherein the similarity between the node attributes is determined according to the following steps: Determining the attribute similarity of a first node between each key equipment node pair or each intermediate equipment node pair according to a first node attribute similarity algorithm, wherein the first node attribute similarity algorithm is determined according to the number of times of operations required by the conversion of the character strings between two key equipment nodes or between attribute values of preset attributes of two intermediate equipment nodes; If the attribute similarity of the first node is smaller than a preset node attribute similarity threshold, determining second node attribute similarity between each key equipment node pair or each intermediate equipment node pair according to a second node attribute similarity algorithm, wherein the second node attribute similarity algorithm is determined according to semantic similarity between two key equipment nodes or between preset attributes of two intermediate equipment nodes and attribute values of the preset attributes; And determining the second node attribute similarity as the similarity between the node attributes.
7. A power knowledge processing apparatus for multi-source data fusion, the apparatus comprising: The system comprises an acquisition unit, a storage unit and a control unit, wherein the acquisition unit is used for acquiring graph structure data corresponding to each preset data source of a target power system, and each graph structure data comprises at least one target equipment node; The mapping unit is used for carrying out functional mapping on each node in each graph structure data to obtain standard function graph structure data corresponding to each preset data source, wherein each node in the standard function graph structure data is a functional node, and the functional node is a node named according to the function; The system comprises a determining unit, a determining unit and a processing unit, wherein the determining unit is used for determining each target equipment node in different preset data sources according to each standard functional diagram structure data and a preset diagram similarity algorithm, and the preset diagram similarity algorithm is used for determining whether the preset equipment node is the target equipment node according to the similarity between network structures and the similarity between node attributes of the preset equipment node in different standard functional diagram structure data; The fusion unit is used for fusing the graph structure data corresponding to each preset data source by taking each target equipment node as a fusion node so as to obtain a multi-data source fusion power knowledge graph; The determining unit is further configured to perform the steps of S310, obtaining an initial similarity matrix, where the initial similarity matrix is n×n dimensions, n is a total number of nodes included in all the graph structure data, the similarity of key elements in the initial similarity matrix is 1, the similarity of non-key elements is 0, the similarity of key elements between the i-th row and i-th column nodes is the similarity of non-key elements, the non-key elements are the similarity between the i-th row nodes and the j-th column nodes, i=1, 2..n, j=1, 2..n, i is not equal to j, S320, determining a current similarity value of each non-key element according to the initial similarity matrix and a preset calculation rule to obtain an intermediate similarity matrix, S330, obtaining an average similarity difference value of the intermediate similarity matrix, S340, determining a current similarity value of each non-key element according to the intermediate similarity matrix and a preset calculation rule to update the intermediate similarity matrix until the average similarity difference value is equal to or greater than a preset difference threshold, determining the intermediate similarity matrix is less than a target similarity matrix according to the preset difference value, and determining the similarity matrix is less than the target similarity matrix.
8. A non-transitory computer readable storage medium having stored therein at least one instruction or at least one program loaded and executed by a processor to implement the method of any one of claims 1-6.
9. An electronic device comprising a processor and the non-transitory computer readable storage medium of claim 8.

Description

Multi-source data fusion power knowledge processing method, device, medium and equipment Technical Field The present application relates to the field of power data processing, and in particular, to a method, an apparatus, a medium, and an electronic device for processing power knowledge with multi-source data fusion. Background In daily operation and management of a power system, a plurality of independent professional management systems are usually deployed to realize fine management and control of different service links. For example, an enterprise asset management system (EAM) mainly focuses on the whole life cycle management of equipment and covers the information of purchase, account setting, maintenance planning and the like of the equipment, a supervisory control and data acquisition System (SCADA) focuses on monitoring the running state of the power equipment in real time and recording dynamic running data such as voltage, current, power and the like, an overhaul work order system focuses on the dispatching, execution and closed-loop management of equipment overhaul tasks and stores flow information such as work order numbers, overhaul contents, processing results and the like, and a Production Management System (PMS), an enterprise resource planning system (ERP) and the like record corresponding data for specific business scenes such as production scheduling, resource allocation and the like. The systems are respectively focused on different aspects of power system operation, and the recorded information types have significant differences, namely, part of the systems are mainly based on static attribute data (such as equipment models and manufacturers in EAM), part of the systems are mainly based on dynamic operation data (such as real-time monitoring values in SCADA), and part of the systems are mainly based on flow data (such as task progress in maintenance work order systems). For key power equipment such as transformers, the related knowledge maintained in each system contains some basic common contents (such as physical positions and core functions of the equipment), but is more embodied as different data contents generated by service scene differences, so that a multi-dimensional information system surrounding the same equipment is formed. However, these service systems often follow respective technical standards and data specifications during construction, and lack of a unified device identification system, so that the same physical device may be given different identifiers or naming manners in different systems, that is, a phenomenon of "synonym" (SAMEENTITY, DIFFERENTNAMES) exists. For example, a certain transformer is numbered "T1023-2020" in the EAM system, labeled "#b12" in the SCADA system, and recorded "main transformer a in the east city area" in the maintenance work order system. The inconsistency of the identification enables equipment data in each system to form an 'information island', so that the fusion and unified management of multi-source data of the same equipment are seriously hindered, the information about static attribute, dynamic operation, maintenance record and the like of the same equipment in different systems cannot be integrated efficiently, advanced applications such as equipment state evaluation, fault diagnosis and the like based on complete data are difficult to support, and the deep mining of the data value of an electric power system and the improvement of business collaborative efficiency are restricted. Disclosure of Invention Aiming at the technical problems, the application provides a power knowledge processing method, a device, a medium and electronic equipment for multi-source data fusion, which at least partially solve the problems in the prior art. In a first aspect of the present application, there is provided a power knowledge processing method for multi-source data fusion, the method comprising the steps of: S100, obtaining graph structure data corresponding to each preset data source of a target power system, wherein each graph structure data comprises at least one target equipment node, and node identifiers of the target equipment nodes contained in each graph structure data are different. And S200, performing function mapping on each node in each graph structure data to obtain standard function graph structure data corresponding to each preset data source, wherein each node in the standard function graph structure data is a function node, and the function node is a node named according to the function. And S300, determining each target equipment node in different preset data sources according to each standard functional diagram structure data and a preset diagram similarity algorithm, wherein the preset diagram similarity algorithm determines whether the preset equipment node is the target equipment node according to the similarity between network structures and the similarity between node attributes of the preset equipment node in different st