This paper presents a novel scbeme for the detection and magnitude determination of turn faults.
Arlificial neural network (ANN) plays a role as fault detector and fault location estimator. This
study is based on simulating the faulty motor on phase (abc) frame. Insensitivity of the technique
to load variation, supply voltage variation and unbalance in supply voltage is Illustrated in this
paper. The results from this study suggests that the proposed method provides a powerful and a
general approach to the ANN based fault detection.