Power dispatching automation system fault tracing method based on information difference graph model

Abstract

The embodiment of the invention provides a power dispatching automation system fault tracing method based on an information difference graph model, and the method comprises the steps: selecting historical data of a power dispatching automation system before and after alarm, obtaining a clustering center through a k-means algorithm, and taking the clustering center as an end point of interval division, wherein the mean value of each interval serves as a discretization result of continuous features; calculating the information entropy of the components of the power dispatching automation systemand the transfer entropy between the components, establishing an information correlation matrix with or without an alarm section, measuring the difference degree before and after the alarm through thechange rate, and obtaining an information difference matrix by adopting a normalization technology; extracting features with relatively high alarm information change of a power dispatching automationsystem and interaction information among the features, further constructing an information difference graph model combining a bidirectional graph and node self-information, and fitting fault degree indexes to sort fault degrees. According to the technical scheme provided by the embodiment of the invention, the fault tracing performance of the power dispatching automation system is improved.

Description

【Technical field】 [0001] The invention relates to a fault tracing method for solving unknown system topology relations in the field of fault location, in particular to a fault tracing method for an electric power dispatching automation system based on an information difference graph model. 【Background technique】 [0002] With the continuous maturity of intelligent technology and network technology, the power dispatching automation system, as a unity of computing and physical processes, is a complex system that integrates computing, communication, and physical environments. The system is usually composed of many components such as servers, storage, network equipment, and application software. Once a component fails, its correlation will affect other components, which will lead to abnormalities or fluctuations in the entire system. Due to the unknown and complex topological relationship between components, it is extremely difficult to trace the fault source. Early traceability...

AI Technical Summary

Problems solved by technology

Modeling-based fault tracing methods mainly include models such as undirected graphs, directed graphs, fault trees, and invariant networks. Among them, the undirected graph model has the problem that the fault propagation direction cannot be determined; the directed graph model is based on complex system logic. Once some components are added or reduced, relevant knowledge in the professional field is required to modify and update, which greatly increases the maintenance cost; the fault tree requires the participation of experts in different fields, and can only be constructed correctly with the help of detailed system knowledge, which cannot meet the requirements for large systems. In addition, due to the subjective factors of expert experience, some causality is ignored, which ultimately limits the deployment of fault trees; the invariant network model is a data-driven way to model the system, which greatly gets rid of the problem of relying heavily on prior knowledge. However, it is necessary to train a huge set of fault candidates to backtrack the root cause fault components

Images