A novel hybrid extreme learning machine-whale optimization algorithm for bearing fault diagnosis

Abstract

In order to avoid fatalities and ensure safe operation, a good and accurate diagnosis method is required. A diagnosis method based on extreme learning machine (ELM) has attracted much attention and the ELM method had been applied in various field of study. The advantages of the ELM method which are rapid learning rate, better generalization performance and ease of implementation makes the ELM method suitable to be used in various field including fault diagnosis fields. However, the performance of the ELM method becomes inefficient due to incorrect selection of neurons number and randomness of input weight and hidden layer bias. Hence, this paper aims to propose a novel hybrid fault diagnosis method based on ELM and whale optimization algorithm (WOA), known as ELM-WOA for bearing fault diagnosis. Four different types of bearing datasets from Case Western Reserve University Bearing Data Centre were used in this paper in order to present the performance of the proposed method. Based on the result, the performance of the proposed method was able to surpass the performance of the conventional ELM.