CN-121502814-B - Intelligent identification management method and system for full life cycle of electric power asset

CN121502814BCN 121502814 BCN121502814 BCN 121502814BCN-121502814-B

Abstract

The invention discloses an intelligent identification management method and system for a full life cycle of an electric power asset, which relate to the technical field of electric power asset management and comprise the steps of generating a unique identification for the electric power asset in a trusted data platform and establishing a file; the method comprises the steps of carrying out embedded modeling on multi-source data under the association of unique identifiers to generate context representation, obtaining causal association results through causal refining, introducing privacy topological factors into the causal association to form privacy enhancement input, generating bearable codes by executing condition countermeasure based on the privacy enhancement input and the anchoring representation, driving dynamic linkage of an operation clock, a verification clock and a warranty clock by utilizing the codes, triggering compliance tasks and writing the results back into files. The balance between privacy protection and topology maintenance is realized, the robustness and the credibility of the identification codes are improved, a closed loop for task linkage and value update is formed in full life cycle management, and the integrity and the compliance of power asset management are improved.

Inventors

LI SHAOXIONG
Zhao haichuan
BAO YINGCHUN

Assignees

内蒙古智通电力设备有限公司

Dates

Publication Date: 20260512
Application Date: 20251124

Claims (7)

1. The intelligent identification management method for the full life cycle of the electric power asset is characterized by comprising the following steps of: Collecting life cycle data of the electric power asset, establishing an archive of the life cycle data of the electric power asset and the unique identifier, and storing the archive in a trusted data platform; under the association of the life cycle data of the electric power asset and the unique identifier, forming context information based on embedded modeling, and obtaining a causal association result through causal refining; The forming of the context information comprises the steps of in a trusted data platform, taking a unique identification of an electric power asset as an index, uniformly aligning design parameters, running state data and maintenance event records, extracting statistical running characteristics according to short-term, medium-term and long-term multi-scale time windows, and generating a context embedding vector reflecting the historical state of equipment and the influence of the event and coupling with a network by combining the time attenuation effect of the maintenance event and the diffusion relation of a power network topology neighborhood; The causal association results obtained through causal refining comprise the steps of calculating causal association relations among all operation features by adopting a causal discovery method constrained by a directed acyclic graph based on the context embedded vector, forming a causal association result matrix, distinguishing direct causal effects from surface correlation, and outputting feature representations after causal propagation; the method comprises the steps of introducing a privacy topological factor in a causal association result to shield sensitive information and keep topological criticality, and forming privacy-enhanced input, wherein the step of introducing the privacy topological factor in the causal association result comprises the steps of setting a shielding coefficient for corresponding dimensions in a causal association result matrix according to a preset sensitive field rule in a trusted data platform, generating a retention coefficient by combining feature importance after causal propagation, and combining the shielding coefficient and the retention coefficient to obtain the privacy topological factor, wherein the privacy topological factor is a set of adjustment parameters set dimension by dimension and used for simultaneously reducing influence of sensitive features in subsequent processing and keeping main causal relation not destroyed; The privacy-enhanced input comprises the steps of correspondingly multiplying the privacy topological factors with feature representations after causal propagation dimension by dimension to obtain context input subjected to privacy protection treatment, weakening sensitive features in each dimension by the context input, and generating privacy-enhanced input data; Generating the trusted identification code comprises the steps of taking the privacy enhanced input as a generation condition in a trusted data platform, combining an anchoring representation bound with a unique identifier, configuring a generation network and a discrimination network, and executing condition countermeasure generation under the combined action of binding constraint, privacy inhibition constraint, structure compliance constraint and stability constraint to obtain the trusted identification code meeting the constraint; And driving the linkage of the running clock, the verification clock and the warranty clock by using the beacon code so as to trigger the compliance task and record the result back to the electric power asset file, thereby realizing value updating.
2. The intelligent identification management method for the whole life cycle of the electric power asset according to claim 1, wherein the life cycle data of the electric power asset comprises design parameters, running states and maintenance records of the electric power asset; Establishing files for the life cycle data of the electric power assets and the unique identifiers comprises the steps of generating the unique identifiers for each electric power asset in a trusted data platform, respectively collecting and archiving design parameters, running states and maintenance records of the electric power assets to corresponding file entries, and binding and storing the unique identifiers in the file entries.
3. The intelligent identification management method for the whole life cycle of the electric power asset according to claim 2, wherein the linkage of driving the operation clock, the verification clock and the warranty clock by using the trusted identification code comprises the steps of extracting operation stress factors and metering deviation information based on the trusted identification code and corresponding context input in a trusted data platform, dynamically accumulating the operation clock according to the operation stress factors, gradually accumulating the verification clock according to the metering deviation information, and decrementing the warranty clock according to a guarantee period in a countdown mode so as to realize the dynamic update of the states of the operation clock, the verification clock and the warranty clock, and keep continuous binding with the trusted identification code.
4. The intelligent identification management method for the whole life cycle of the power asset, as set forth in claim 3, wherein triggering the compliance task and recording the result back to the power asset file comprises generating a trigger instruction when the state of any clock reaches a preset threshold condition, executing task scheduling in a trusted data platform according to the compliance priority based on the trigger instruction to form a corresponding operation and maintenance work order, verification task or warranty processing order, binding and storing the operation record with the trusted identification code after the task is completed, and updating state information and value information in the power asset file to realize closed-loop management of the whole life cycle.
5. An intelligent identification management system for a full life cycle of an electric power asset is based on the intelligent identification management method for the full life cycle of the electric power asset according to any one of claims 1 to 4, and is characterized in that: The data acquisition module is used for acquiring design parameters, running states and maintenance records of the electric power asset, establishing files with the unique identification in the trusted data platform and performing binding storage; The context modeling module is used for uniformly aligning the design parameters, the running state data and the maintenance event records in the trusted data platform by taking the unique identification as an index, and generating a context embedding vector; The causal refining module is used for obtaining a causal association result matrix and outputting a causal propagated characteristic representation by adopting a causal discovery method of directed acyclic graph constraint based on the context embedded vector; The privacy topological factor construction module is used for generating privacy topological factors in the trusted data platform according to preset sensitive field rules and combining the privacy topological factors with feature representations after causal transmission to form privacy enhanced input; the countermeasure generation module is used for taking the privacy enhanced input as a generation condition in the trusted data platform, combining the anchored representation bound with the unique identifier, executing the condition countermeasure generation, obtaining a trusted identifier code and binding and storing the trusted identifier code and the asset file; The clock linkage module is used for driving the running clock, the verification clock and the maintenance clock to dynamically update based on the state of the identifiable code, triggering a compliance task when a threshold condition is reached, and recording the result back to the electric power asset file to realize value update.
6. The computer equipment comprises a memory and a processor, wherein the memory stores a computer program, and the computer equipment is characterized in that the processor realizes the steps of the intelligent identification management method facing the full life cycle of the electric power asset according to any one of claims 1-4 when executing the computer program.
7. A computer readable storage medium, on which a computer program is stored, is characterized in that the computer program, when executed by a processor, implements the steps of the intelligent identification management method for a full life cycle of an electric power asset according to any one of claims 1 to 4.

Description

Intelligent identification management method and system for full life cycle of electric power asset Technical Field The invention relates to the technical field of electric power asset management, in particular to an intelligent identification management method and system for the whole life cycle of an electric power asset. Background The electric power asset is used as a core resource for the operation of the electric power system, and the full life cycle management of the electric power asset is directly related to the power supply safety, the operation efficiency and the economic benefit. Conventional power asset management models typically rely on independent information systems for each business division, such as material procurement systems, operation monitoring systems, certification systems, financial systems, and the like. The lack of a unified data index and a communication mechanism between the systems causes the problems of repeated input, inconsistent caliber and information island of asset information in different links. For example, the state parameters collected by the equipment in the operation stage are difficult to trace back to the design file in time, and verification results and maintenance records are often disjointed with the operation state data, so that the accuracy of asset life assessment and operation and maintenance decision is affected. In recent years, with the development of new technologies such as internet of things sensing, blockchain, artificial intelligence and the like, the full life cycle management of electric power assets gradually evolves to an intelligent and digital direction. The sensor of the Internet of things can collect operation data such as voltage, current and temperature in real time, the blockchain technology can realize tamper-proof evidence storage of data shared by multiple parties, and the artificial intelligence algorithm is also applied to health assessment and life prediction. However, the existing method still has the following technical defects that firstly, a unified asset identification management mechanism is lacked, equipment codes in different systems are difficult to match, so that complete files cannot be formed on data, secondly, most of existing state modeling depends on surface correlation analysis, real causal relations among variables are difficult to identify, so that misjudgment is easy to occur in prediction and diagnosis, thirdly, contradiction exists between privacy protection and data sharing, key parameters (such as geographic positions and verification records) are needed to participate in analysis, leakage must be prevented, fourthly, the existing intelligent identification generation method is mostly in a single hash or coding mode, lacks robustness resistance and is easy to forge or tamper, and fifthly, on asset compliance management, the operation period, verification period and warranty period are often scattered to manage, and a unified linkage mechanism is lacked, so that task scheduling is lagged or repeated, and a closed loop with updated value cannot be formed. Disclosure of Invention The present invention has been made in view of the above-described problems occurring in the prior art. Therefore, the intelligent identification management method for the full life cycle of the electric power asset solves the problems that the existing full life cycle management of the electric power asset has serious information island, insufficient causality identification and difficult closed loop formation in the aspects of data penetration, causality modeling and compliance linkage, and the decentralized management of operation, verification and warranty periods is difficult. In order to solve the technical problems, the invention provides the following technical scheme: In a first aspect, the present invention provides an intelligent identification management method for a full life cycle of an electric power asset, which includes collecting life cycle data of the electric power asset, creating a file of the life cycle data of the electric power asset and a unique identification, and storing the file in a trusted data platform; under the association of the life cycle data of the electric power asset and the unique identifier, forming context information based on embedded modeling, and obtaining a causal association result through causal refining; Introducing a privacy topological factor into the causal association result to shield sensitive information and keep topological criticality, so as to form privacy-enhanced input; Performing countermeasure generation on the privacy-enhanced input to obtain a robust trusted identification code; And driving the linkage of the running clock, the verification clock and the warranty clock by using the beacon code so as to trigger the compliance task and record the result back to the electric power asset file, thereby realizing value updating. As a preferable scheme of the intelligent identif