Approach for Fault Diagnosis with Improved Compensating Fuzzy Neural Network

摘要

In order to solve the problem that the reliability of fault diagnosis with standard thresholds is low in some conditions, this paper presents a approach of fault diagnose with improved compensating fuzzy neural network (ICFNN). The simulation of fault diagnosis based on a certain missile control system proves that with 6.67% relative reduction of the parameter of the rate channel under the condition of a stable system this method can correctly diagnose and fault diagnosis with standard thresholds cannot. The fault diagnosis by ICFNN is better than that by standard signal in reliability.