Implementation of Efficient Artificial Neural Network Data Fusion Classification Technique for Induction Motor Fault Detection

Abstract

Reliability measurement and estimation of an industrial system is a difficult and essential problematic task for control engineers. In this context reliability can be described as the probability that machine network will implement its proposed functions under the observing condition throughout a specified time period of running machine system network. In this study single sensor method is applied for fault diagnosis depending on identification of single parameter. At early stages it is hard to diagnose machine fault due to ambiguities in modeling environment. Due to these uncertainties and ambiguities in modeling, decision making become difficult and lead to high financial loss. To overcome these issues between the machine fault symptoms and estimating the severity of the fault; a new method of artificial intelligence fault diagnosis based approach Dempster–Shafer theory has been proposed in this paper. This theory will help in making accurate decision of the machine condition by fusing information from different sensors. The experimental results demonstrate the efficient performance of this theory which can be easily compared between unsurpassed discrete classifiers with the single sensor source data.

Keywords

• Dempster–Shafer theory
• data fusion
• fault diagnosis
• artificial neural network
• fast Fourier transform