Fault diagnosis of flying control system servo actuator based on Elman neural network

Abstract

Servo actuator is one most important section of flying control system, and faults also often happen to it. The fault detection and diagnosis technology is seriously important to improve the reliability of servo actuator. The paper proposes a method of fault diagnosis based on Elman neural network, using its non-linear distributed processing and dynamic feature reflecting ability to detect servo actuator fault. Then, neural network algorithm is applied to simulation. The result indicated that the method could accurately identify the servo actuator fault. Meanwhile, compared with BP neural network, the advantage of Elman neural network in fault diagnosis is confirmed.