Fault Simulation and Diagnosis of Vacuum System Based on a Simulation Platform

Abstract

The simulation model of vacuum system is an important part of the simulation model of thermal power unit, of which dynamic and static characteristics are of direct effect on the simulating efficacy of the model. However, fault simulation and diagnosis technology is mainly applied to the main equipment and systems instead of some important auxiliary systems as the vacuum system. In this thesis, considering the non-linear characteristic of faults in vacuum system and strength of BP neural network in learning, BP neural network was applied to fault diagnosis in fault simulation model of vacuum system and two factual faults diagnosis in vacuum system. The application of the fault diagnosis technology based on simulation platform provides ground work for setting up fault detection and diagnosis system on vacuum system in power plant DCS.