Fault Diagnosis of Aero-engine Gas Path Base on PSO-SVM

Abstract

Aiming at the high incidence of faults and high maintenance cost of aero-engine gas path components, this paper adopts the condition-based maintenance mode and introduces the fault diagnosis method combining particle swarm optimization (PSO)with support vector machine(SVM). Firstly, the fault diagnosis method of aero-engine gas path based on SVM is proposed. Then, kernel function parameters and penalty coefficients of SVM are optimized by PSO. Finally, the aero-engine gas path fault diagnosis model based on PSO-SVM is established. The simulation results show that the design method of fault diagnosis of aero-engine Gas path based on PSO-SVM has advantages of short prediction time, high prediction efficiency and higher accuracy than the traditional diagnosis method based on SVM, reached 100%. Moreover, the effect of PSO on kernel function parameters and penalty coefficients is better than that of other optimization algorithms.