Fault feature extraction for multiple electrical faults of aviation electro-mechanical actuator based on symbolic dynamics entropy

In this paper, a simple fault feature extraction method based on symbolic dynamics entropy is proposed for multiple electrical faults diagnosis in an electro-mechanical actuator. The symbolic dynamics feature is calculated from a time series, which is extracted from the dc-link current of the actuator. Comparing with other fault feature extraction scheme, this method has some advantages such as simpler computation, less data requirement and it can realize multiple faults detection. The proposed method is verified through simulation experiment in the first stage of our study. Fault injection methods on four typical electrical faults of the actuator are studied firstly. Then, by doing experiment under specific position command and load condition, the effectiveness of the fault feature to detect multiple electrical faults is verified.