Distributed signature analysis of induction motors using Artificial Neural Networks

Motor current signature analysis is a modern approach to fault diagnose and classification for induction motors. Many studies reported successful implementation of MCSA in laboratory situations whereas the method was not so successful in real industrial situation due to propagation of neighbor faults and unwanted noise signals. This paper investigate the correlation between different observations of events in order to provide a more accurate estimation of behavior of electrical motors at a given site. An analytical framework has been implemented to correlate and classify independent fault observations and diagnose the type and identify the origin of fault symptoms. The fault diagnosis algorithm has two layers. Initially outputs of all sensors are processed to generate fault indicators. These fault indicators then are to be classified using an Artificial Neural Network. A typical industrial site is taken as a case study and simulated to evaluate the concept of distributed fault analysis.