Predicting the remaining useful life of the product based on vibration signals and genetic algorithm

Khakbazan, Mojtaba; Ramezani, Saeid; Yousefi Tezerjan, Mustafa

doi:JABM.3.2.11235564.35254154

Predicting the remaining useful life of the product based on vibration signals and genetic algorithm

Document Type : The scientific research paper

Authors

Mojtaba Khakbazan ¹

Saeid Ramezani ²

Mustafa Yousefi Tezerjan ³

¹ Ph.D. student of industrial engineering at Ivanki University

² Member of the Faculty of Industrial Engineering, Faculty of Technology and Engineering, Imam Hossein University (AS), Tehran, Iran

³ Member of the Faculty of Applied Science University, Alborz, Iran

JABM.3.2.11235564.35254154

Abstract

In today's competitive world, maintaining product quality and accurately predicting its useful life is of particular importance. Sudden failures such as bearing failures can have significant financial and time effects on production. In order to deal with these challenges, this research presents an intelligent model based on vibration analysis and genetic algorithms, which is able to predict the remaining useful life of bearings with high accuracy. The proposed model consists of four parts: the quality indicators prediction part, the initial state of deterioration part, the information integration part and the optimization part. This model not only helps improve maintenance strategies, but also leads to increased productivity and profitability of companies by reducing unexpected downtime periods. Experimental results show that the performance of our model is better than conventional methods such as autoregression models, especially thanks to the use of optimized parameters in different steps.

Keywords

Life prediction

genetic algorithm

data mining

product marketing

Subjects