Abstract
The reliability and efficiency of 3-phase induction motors are crucial
for various industrial applications. However, the timely fault detection and
prediction in these motors remains a significant challenge. Traditional fault
prediction methods cannot often effectively handle complex fault patterns and
dynamic operating conditions. Due to this, researchers have focused towards
soft computing techniques for enhanced fault prediction accuracy. This paper
reviews the principles and methodologies of traditional fault prediction
methods, highlighting their limitations in accurately predicting faults in 3-phase
induction motors. This research paper explores the transition from conventional
fault methods to soft computing techniques such as fuzzy inference systems,
artificial neural networks and adaptive neuro-fuzzy inference systems. These
techniques enable adaptive learning, helping the system to recognise and adapt
to complex fault patterns and varying operating conditions. The results show
the detailed analysis of eight types of crucial faults using various computing
techniques under different parametric conditions.