An adaptive network based fuzzy diagnostic system for linear induction motor drives

The working conditions identification of technical systems is the challenge of technicians who deal with diagnostic problems. Sometimes, there is not a mathematical model which is able to describe the behaviour of the system or if there is its complexity do not allow a useful implementation to perform an on-line and real-time diagnostic process. In such cases the use of diagnostic techniques based on artificial intelligence is suitable. The aim of this paper is to present an adaptive-network-based fuzzy diagnostic system in which adaptive networks are used to construct as symptoms-faults mapping for linear induction motor drives. Such mapping is carried out by means of learning procedure based on experimental data measured in several normal and faulty working conditions of the system under diagnosis. The proposed diagnostic system has been experimentally validated through plenty of tests