In this paper, we propose a Negative Selection Algorithm (NSA)-based motor fault detection system. Only the feature signals of the healthy motors are needed here for generating the NSA detectors. Different from the conventional fault detection approaches, no prior knowledge of the motor fault types is assumed to be known beforehand in the proposed scheme. Our motor fault detection method is examined using both the rotor and stator faults in computer simulations. We further explore its applicability in case of fault detection with varying motor loads.
Mendeley saves you time finding and organizing research
Choose a citation style from the tabs below