A diagnostic method for faulty equipment in large power grids and backbone optical communication systems
By establishing an operational fault correlation model in the large power grid and backbone optical communication system, and using knowledge graphs and preset rule trees for fault reasoning, the problem of difficult fault location in the power backbone network was solved, and rapid and accurate fault location and operation and maintenance optimization were achieved.
Patent Information
- Application Number
- CN202311473139.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-11-07
- Publication Date
- 2026-11-17
- Estimated Expiration
- 2043-11-07
AI Technical Summary
Existing technologies cannot quickly and accurately locate faults in the power backbone network, resulting in a waste of human and material resources in operation and maintenance. Furthermore, the interfaces between different disciplines are unclear, making it difficult to achieve intensive and standardized management.
A fault association model between a large power grid and a backbone optical communication system is established using knowledge graph technology. By learning and labeling fault candidate relationship entities, a knowledge connection network is formed. Combined with a preset rule tree, reasoning is performed to identify faulty equipment and generate fault handling reports.
It enables rapid and accurate location of faulty equipment, improves fault retrieval efficiency, optimizes communication operation, and enhances the power grid's emergency response capabilities and operation and maintenance efficiency.
Smart Images

Figure CN117579143B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of power communication system technology, and more specifically, to a method for diagnosing faults in large power grid and backbone optical communication equipment. Background Technology
[0002] The power backbone network, as the second physical network coexisting with the power grid, is a crucial support for the safe, stable, and economical operation of the power grid. The power backbone network can be divided into transmission networks, data networks, and support networks. Each network contains a large number of communication and transmission equipment, such as PCM (Power Transceiver Machines) for access, SDH (Single-Directional Hierarchy) for transmission, data communication network equipment, dispatch data network equipment, and communication resources such as optical cables and power supplies. The communication equipment comes from numerous manufacturers and involves complex equipment types, making it impossible to establish a complete end-to-end cause analysis when a fault occurs. Each network has its own independent network management and monitoring system, and various monitoring and equipment management data are scattered across different business systems such as communication management systems and dispatch automation systems. The ledgers and historical fault records for the same equipment in different business systems lack a unified format, making interoperability difficult and hindering accurate fault location.
[0003] Furthermore, the interfaces between communication, operation, maintenance, and protection specialties in the existing backbone network operation and maintenance system are relatively vague. In the actual fault handling process, due to the inability to accurately locate the cause of the fault, multiple specialties often have to go to the site to investigate the cause of the problem, resulting in a waste of human and material resources and restricting the intensive, standardized, and lean management of backbone communication network operation and maintenance. Summary of the Invention
[0004] The purpose of this invention is to provide a diagnostic method for faulty equipment in large power grids and backbone optical communication systems, so as to solve the current technical problem of being unable to quickly and accurately locate faults in the power backbone network.
[0005] To achieve the above objectives, the technical solution of the present invention is: a method for diagnosing faults in large power grid and backbone optical communication equipment, comprising the following steps: S1. Acquire the first power grid operation data every predetermined first cycle, first preprocess the data, then perform entity identification on the cleaned data, then associate fault entities, and finally output a fault association set that satisfies the specified relationship to achieve knowledge extraction of operation fault association. S2. Acquire the second power grid operation data every predetermined second cycle; S3. Based on the first power grid operation data and the second power grid operation data, determine and analyze whether a power grid fault has occurred. If so, perform power grid topology coloring analysis based on the first power grid operation data and the second power grid operation data respectively. Using knowledge graph technology, establish an operation fault association model between the large power grid and the backbone optical communication system. This model expresses the possible fault association relationships between the large power grid and the backbone optical communication system in the form of a graph. Using a fault candidate relationship entity pair learning and labeling method based on weak supervision, establish relationship entity pairs to form a knowledge connection network. Finally, establish a knowledge graph of operation fault relationships between the large power grid and the backbone optical communication system to obtain a list of possible faulty equipment. S4. Traverse the list of potentially faulty devices and infer the faulty device based on the operating information of the potentially faulty device according to the matching preset rule tree. Specifically, this includes: taking a potentially faulty device from the list for inference; matching the corresponding rule tree based on the device type, voltage level, and wiring method of the inferred potentially faulty device; starting from the inferred node in the matching rule tree, obtaining the child nodes under the inferred node; determining whether the rule of at least one node under the inferred node corresponds to the operating information of the inferred potentially faulty device; if so, incrementing the confidence of the inferred potentially faulty device by 1 and storing the corresponding node in the queue to be inferred; where the corresponding node is the child node under the inferred node whose rule corresponds to the operating information of the inferred potentially faulty device; taking a node from the queue to be inferred for inference and determining whether the node is a... Leaf nodes, by assuming a certain fault exists in the corresponding equipment, clean, parse, classify, and complete the real-time and historical fault data of the equipment. If a new fault type appears, it is marked. After the collected data is processed and forwarded to the data storage area, it is stored in distributed storage, time-series database, real-time cache, or relational database according to the data usage requirements, and is also saved in the fault database. Subsequently, during the operation of the power grid, real-time operation status data is collected. The operation status data is used as the data to be cleaned and matched with each fault expression in the fault database. Data that does not match is cleaned up, and data that matches successfully is filtered out as valid fault data. Finally, the knowledge graph of the relationship between the power grid and the backbone optical communication system operation faults is used to perform fault location analysis on the valid fault data to identify possible equipment faults. S5. Based on the identified equipment fault information and in conjunction with the dispatch fault handling management regulations, fault handling plans, dispatch procedures and anomaly monitoring manual, generate a fault handling report and send it to relevant personnel so that the fault can be handled in a timely manner.
[0006] In one embodiment of the present invention, the fault expression is generated as follows: based on the power grid, fault type and fault conditions, fault rules are constructed, fault expressions are generated based on the fault rules, and the fault expressions are stored in the fault database.
[0007] In one embodiment of the present invention, both the first power grid operation data and the second power grid operation data include power data and optical communication data; the power data includes parameters of power grid voltage, current, and frequency, and the optical communication data includes parameters of optical signal intensity, wavelength, and transmission rate; the duration of the first cycle is longer than the duration of the second cycle.
[0008] In one embodiment of the present invention, the knowledge graph technology includes graph databases and graph neural networks. By utilizing these technologies, useful information can be extracted from a large amount of operational data, and an operational fault correlation model between the power grid and the backbone optical communication system can be established.
[0009] In one embodiment of the present invention, the operating data of the potentially faulty device is the power data and optical communication data related to the potentially faulty device in the second power grid operating data.
[0010] In one embodiment of the present invention, the preset rule tree defines the association between power grid equipment information and communication equipment information when different types, voltage levels and wiring methods of equipment fail; the traversal of the list of possible faulty equipment is to infer the faulty equipment based on the operating information of the possible faulty equipment according to the matching preset rule tree.
[0011] In one embodiment of the present invention, the fault location analysis includes using an association model for deep learning, pattern recognition, and anomaly detection to identify possible equipment faults.
[0012] In one embodiment of the present invention, the fault report includes information on fault type, fault location, and fault severity, and can be displayed in the form of charts and text to facilitate relevant personnel to quickly understand and promptly handle the fault.
[0013] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention proposes a diagnostic method for faulty equipment in large power grids and backbone optical communication systems. Based on a knowledge graph of operational fault associations between large power grids and backbone optical communication systems, this method quickly and accurately determines the location and type of faulty equipment by identifying and matching fault patterns. This method changes the existing method of power grid fault information retrieval, achieving fault retrieval through reasoning and displaying fault information to power operation and maintenance personnel in a graphical manner, greatly improving fault retrieval efficiency.
[0014] 2. Utilize textual data such as dispatch fault handling management regulations, fault handling plans, dispatch procedures, and anomaly monitoring manuals, as well as structured data exported from business systems, to achieve fault information parsing, fault judgment, and fault handling through knowledge graph technology. This helps business personnel quickly analyze the causes of accidents, comprehensively grasp key information for fault handling, and make auxiliary decisions to improve the emergency response capabilities of the power grid.
[0015] 3. By integrating data from multiple disciplines such as communication measurement, relay protection, and dispatch data network, and combining this year's fault information, we analyze the weaknesses of equipment and optical paths, optimize communication operation modes, and realize the transformation from passive emergency repair to proactive operation and maintenance. Attached Figure Description
[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.
[0017] Figure 1 A flowchart illustrating a diagnostic method for faulty equipment in a large power grid and backbone optical communication system, provided in this application embodiment; Figure 2 A fault reasoning model diagram for a diagnostic method for faulty equipment in a large power grid and backbone optical communication provided in this application embodiment; Figure 3 This application provides a fault analysis knowledge graph search engine architecture diagram for a diagnostic method of faulty equipment in a large power grid and backbone optical communication system. Detailed Implementation
[0018] The technical solution of the present invention will now be described in detail with reference to the accompanying drawings.
[0019] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.
[0020] This invention provides a method for diagnosing faults in large power grids and backbone optical communication equipment, comprising the following steps: S1. Acquire the first power grid operation data every predetermined first cycle, first preprocess the data, then perform entity identification on the cleaned data, then associate fault entities, and finally output a fault association set that satisfies the specified relationship to achieve knowledge extraction of operation fault association. S2. Acquire the second power grid operation data every predetermined second cycle; S3. Based on the first power grid operation data and the second power grid operation data, determine and analyze whether a power grid fault has occurred. If so, perform power grid topology coloring analysis based on the first power grid operation data and the second power grid operation data respectively. Using knowledge graph technology, establish an operation fault association model between the large power grid and the backbone optical communication system. This model expresses the possible fault association relationships between the large power grid and the backbone optical communication system in the form of a graph. Using a fault candidate relationship entity pair learning and labeling method based on weak supervision, establish relationship entity pairs to form a knowledge connection network. Finally, establish a knowledge graph of operation fault relationships between the large power grid and the backbone optical communication system to obtain a list of possible faulty equipment. S4. Traverse the list of potentially faulty devices and use the operating information of these devices to infer the faulty device based on a pre-defined rule tree. Specifically, this includes: retrieving a potentially faulty device from the list for inference; matching the device type, voltage level, and wiring method of the inferred potentially faulty device against the corresponding rule tree; starting from the node in the matched rule tree, obtaining the child nodes under the inferred node; determining whether the rule of at least one of the child nodes corresponds to the operating information of the inferred potentially faulty device; if so, incrementing the confidence level of the inferred potentially faulty device by 1 and storing the corresponding node in the inference queue; where the corresponding node is the child node under the inferred node whose rule corresponds to the operating information of the inferred potentially faulty device; retrieving a node from the inference queue for inference; determining whether the node is a leaf node; and assuming that the corresponding device has a certain fault, analyzing the device's real-time data and historical fault data. The fault data is cleaned, parsed, classified, and completed. New fault types are marked if they appear. After processing, the collected data is forwarded to the data storage area and stored in distributed storage, time-series databases, real-time caches, or relational databases according to usage requirements. This data is also stored in the fault database. Subsequently, during power grid operation, real-time operating status data is collected and used as cleaned data. This data is matched against fault expressions in the fault database. Unmatched data is removed, and successfully matched data is selected as valid fault data. Fault expressions are generated as follows: fault rules are constructed based on the power grid, fault type, and fault conditions; fault expressions are generated based on these rules and stored in the fault database. Finally, a knowledge graph of the relationship between the power grid and backbone optical communication system's operating faults is used to perform fault location analysis on the valid fault data, identifying potential equipment faults. S5. Based on the identified equipment fault information and in conjunction with the dispatch fault handling management regulations, fault handling plans, dispatch procedures and anomaly monitoring manual, generate a fault handling report and send it to relevant personnel so that the fault can be handled in a timely manner.
[0021] In this embodiment of the invention, both the first power grid operation data and the second power grid operation data include power data and optical communication data; the power data includes parameters such as voltage, current, and frequency of the power grid, and the optical communication data includes parameters such as intensity, wavelength, and transmission rate of the optical signal; the duration of the first cycle is longer than the duration of the second cycle.
[0022] In this embodiment of the invention, the knowledge graph technology includes graph databases and graph neural networks. By utilizing these technologies, useful information can be extracted from a large amount of operational data, and an operational fault correlation model between the power grid and the backbone optical communication system can be established. The operational data of the potentially faulty equipment consists of power data and optical communication data related to the potentially faulty equipment in the second power grid operational data. The preset rule tree defines the correlation between power grid equipment information and communication equipment information when equipment of different types, voltage levels, and wiring methods fails; traversing the list of potentially faulty equipment involves inferring the faulty equipment based on the operational information of the potentially faulty equipment according to the matching preset rule tree.
[0023] In this embodiment of the invention, the fault location analysis includes using an association model for deep learning, pattern recognition, and anomaly detection to identify possible equipment faults.
[0024] In this embodiment of the invention, the fault report includes information on fault type, fault location, and fault severity, and can be displayed in the form of charts and text to facilitate relevant personnel to quickly understand and promptly handle the fault.
[0025] The above are preferred embodiments of the present invention. Any changes made to the technical solution of the present invention that do not exceed the scope of the technical solution of the present invention shall fall within the protection scope of the present invention.
Claims
1. A method for diagnosing faults in large power grid and backbone optical communication equipment, characterized in that, Includes the following steps: S1. Acquire the first power grid operation data every predetermined first cycle, first preprocess the data, then perform entity identification on the cleaned data, then associate fault entities, and finally output a fault association set that satisfies the specified relationship to achieve knowledge extraction of operation fault association. S2. Acquire the second power grid operation data every predetermined second cycle; S3. Based on the first power grid operation data and the second power grid operation data, determine and analyze whether a power grid fault has occurred; If so, then based on the first power grid operation data and the second power grid operation data, a power grid topology coloring analysis is performed respectively. Using knowledge graph technology, an operation fault association model between the large power grid and the backbone optical communication system is established. This model expresses the possible fault association relationships between the large power grid and the backbone optical communication system in the form of a graph. A fault candidate relation entity pair learning and labeling method based on weak supervision is adopted to establish relation entity pairs and form a knowledge connection network. Finally, an operation fault relationship knowledge graph between the large power grid and the backbone optical communication system is established to obtain a list of possible faulty equipment. S4. Traverse the list of potentially faulty devices and infer the faulty device based on the operating information of the potentially faulty device according to the matching preset rule tree. Specifically, this includes: taking a potentially faulty device from the list for inference; matching the corresponding rule tree based on the device type, voltage level, and wiring method of the inferred potentially faulty device; starting from the inferred node in the matching rule tree, obtaining the child nodes under the inferred node; determining whether the rule of at least one node under the inferred node corresponds to the operating information of the inferred potentially faulty device; if so, incrementing the confidence of the inferred potentially faulty device by 1 and storing the corresponding node in the queue to be inferred; where the corresponding node is the child node under the inferred node whose rule corresponds to the operating information of the inferred potentially faulty device; taking a node from the queue to be inferred for inference and determining whether the node is a... Leaf nodes, by assuming a certain fault exists in the corresponding equipment, clean, parse, classify, and complete the real-time and historical fault data of the equipment. If a new fault type appears, it is marked. After the collected data is processed and forwarded to the data storage area, it is stored in distributed storage, time-series database, real-time cache, or relational database according to the data usage requirements, and is also saved in the fault database. Subsequently, during the operation of the power grid, real-time operation status data is collected. The operation status data is used as the data to be cleaned and matched with each fault expression in the fault database. Data that does not match is cleaned up, and data that matches successfully is filtered out as valid fault data. Finally, the knowledge graph of the relationship between the power grid and the backbone optical communication system operation faults is used to perform fault location analysis on the valid fault data to identify possible equipment faults. S5. Based on the identified equipment fault information and in conjunction with the dispatch fault handling management regulations, fault handling plans, dispatch procedures and anomaly monitoring manual, generate a fault handling report and send it to relevant personnel so that the fault can be handled in a timely manner.
2. The diagnostic method for faulty equipment in a large power grid and backbone optical communication system according to claim 1, characterized in that, The fault expression generation method is as follows: based on the power grid, fault type and fault conditions, fault rules are constructed, fault expressions are generated according to the fault rules, and the fault expressions are saved in the fault database.
3. The diagnostic method for faulty equipment in a large power grid and backbone optical communication system according to claim 1, characterized in that, Both the first power grid operation data and the second power grid operation data include power data and optical communication data; the power data includes parameters such as voltage, current, and frequency of the power grid, and the optical communication data includes parameters such as intensity, wavelength, and transmission rate of the optical signal; the duration of the first cycle is longer than the duration of the second cycle.
4. The diagnostic method for faulty equipment in a large power grid and backbone optical communication system according to claim 1, characterized in that, The knowledge graph technology includes graph databases and graph neural networks. By utilizing these technologies, useful information can be extracted from a large amount of operational data, and an operational fault correlation model between the power grid and the backbone optical communication system can be established.
5. The diagnostic method for faulty equipment in a large power grid and backbone optical communication system according to claim 1, characterized in that, The operating data of the potentially faulty device is the power data and optical communication data related to the potentially faulty device in the second power grid operating data.
6. The diagnostic method for faulty equipment in a large power grid and backbone optical communication system according to claim 1, characterized in that, The preset rule tree defines the association between power grid equipment information and communication equipment information when different types, voltage levels, and wiring methods of equipment fail; the traversal of the list of possible faulty devices involves inferring the faulty devices based on the operating information of the possible faulty devices according to the matching preset rule tree.
7. The diagnostic method for faulty equipment in a large power grid and backbone optical communication system according to claim 1, characterized in that, The fault location analysis includes using association models for deep learning, pattern recognition, and anomaly detection to identify potential equipment faults.
8. The diagnostic method for faulty equipment in a large power grid and backbone optical communication system according to claim 1, characterized in that, The fault handling report includes information on the fault type, fault location, and fault severity, and can be displayed in the form of charts and text to facilitate relevant personnel to quickly understand and handle the fault in a timely manner.
Citation Information
Patent Citations
Backbone optical communication system knowledge extraction method and system
CN116956919A
Fault processing method and apparatus, device, and storage medium
WO2023011160A1
