Research On Fault Detection And Diagnosis Of Control System For Thermal Power Plant Based On Multi-Agent

Fault diagnostic technology is essential to ensure the operation of thermal power plant's reliability and safety. According to the characteristic of control system in thermal power plant, theories and approaches of Fault Detectiong and Diagnosis (FDD) are mainly studied in this paper. Firstly, from the point of component of the system, a novel fault detection algorithm for actuator and sensor using trend analysis of input and output signals is proposed. Secondly, the paper introduces Agent technology to design the superheated steam control system fault detection and diagnosis architecture, using morphology-wavelet analysis as a diagnostic tool for fault. Thirdly, simulation results on superheated steam control system show its good performance in identifying the fault occurrence and diagnosing the fault style in different operation points from the strong noise.