Research On Gas Path Fault Diagnosis For Aero-Engine

As the complex structure and poor working conditions of aero-engine, maintenance has become the main task of aircraft maintenance. People had developed a technique which named condition maintenance for higher utilization and maintenance efficiency. And condition monitoring and fault diagnosing are the basic technologies which to achieve the purpose of condition maintenance.This article takes the aircraft engine as the object of study, and analyzes gas-path fault of air engines.Firstly, the paper summarizes diagnostic techniques of gas-path fault, analyzes the types of gas-path fault, the basic failure mechanism and principles of gas path fault. And establishes a simulation model of JT9D, which is a big ratio turbine-fan by using GSP for engine simulation.Secondly, one way to improve the standard self-adaptive neural fuzzy inference system (IANFIS) by using Fletcher-Reeves Update conjugates gradient method and proportion conjugates gradient method. It helps to solve the problems that the training speed is too slow and easily falls into local minimum as a result of using BP training algorithm. Then it should get data through the simulation model of the JT9D engine which has established before, select the high-and low rotor speed deviation of conversion, exhaust temperature difference and fuel flow deviation as the fault features. And analyze the IANFIS Algorithm in MTALAB. It shows us that the improved algorithm has achieved good results on training time and the diagnosis rate for comparative analyzing of both methods.At last, a simple gas path fault diagnosis software is developed by using VC++6.0 and MATLAB.