Electromechanical Actuator Modeling and Its Application in Fault Diagnosis

As electromechanical actuators (EMA) play an increasingly important role in flight control systems, effective fault diagnosis for them has become an important subject. Considering the actual situation and the cost of setting up the fault, this paper establishes the simulation model of EMA, and realizes the fault diagnosis using neural network: firstly, the simulation model of EMA is built in AMESim, and the characteristic parameters which can represent the working state of EMA are selected, and the original fault data are obtained by fault injection. Then wavelet analysis is used to extract fault features to train BP neural network. The results show that the diagnostic accuracy is more than 95% when the parameters are set reasonably, and the validity of the method is verified.