A Novel Fault Diagnosis of Induction Motor by Using Various Soft Computation Techniques: BESO-RDFA

Abstract

This paper presents a hybrid prediction technique for fault detection of induction machines. The established hybrid forecast scheme signifies the combined execution of Bald-Eagle- Search-Optimization (BESO) and Random-Decision-Forest-Algorithm (RDFA), called as BESO-RDFA prediction scheme. This proposed technique is used to predict the fault within a short period in the rotating machines. By considering the machine defects the RDFA is trained by using the BESO-based exact prediction with data in online mode. The MATLAB/Simulink work platform is employed to execute the model, which is then assessed using multiple techniques to forecast attributes and models of impending stator failure. A new robust diagnostic design is established to analyze the incipient stator winding failures. Simulation analysis shows the detection and isolation method with great sensitivity indicating the incipient winding failures.