Rolling Bearing Fault Diagnosis Based on Deep Learning and Autoencoder Information Fusion

Jianpeng Ma,Chengwei Li,Guangzhu Zhang

doi:10.3390/sym14010013

Jianpeng Ma, Chengwei Li + Show 1 more

Open Access

https://doi.org/10.3390/sym14010013

Copy DOI

Abstract

The multisource information fusion technique is currently one of the common methods for rolling bearing fault diagnosis. However, the current research rarely fuses information from the data of different sensors. At the same time, the dispersion itself in the VAE method has asymmetric characteristics, which can enhance the robustness of the system. Therefore, in this paper, the information fusion method of the variational autoencoder (VAE) and random forest (RF) methods are targeted for subsequent lifetime evolution analysis. This fusion method achieves, for the first time, the simultaneous monitoring of acceleration signals, weak magnetic signals and temperature signals of rolling bearings, thus improving the fault diagnosis capability and laying the foundation for subsequent life evolution analysis and the study of the fault–slip correlation. Drawing on the experimental procedure of the CWRU’s rolling bearing dataset, the proposed VAERF technique was evaluated by conducting inner ring fault diagnosis experiments on the experimental platform of the self-research project. The proposed method exhibits the best performance compared to other point-to-point algorithms, achieving a classification rate of 98.19%. The comparison results further demonstrate that the deep learning fusion of weak magnetic and vibration signals can improve the fault diagnosis of rolling bearings.

Highlights

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations
Damage to rolling bearings may lead to the complete failure of mechanical equipment and even cause economic loss including human casualties [3,4]
The condition of a rolling bearing can be determined by measuring vibration [5], acoustic [6], thermal [7] and oil-based [8] signals

Summary

Introduction

The condition of a rolling bearing can be determined by measuring vibration [5], acoustic [6], thermal [7] and oil-based [8] signals. The electromagnetic induction detection [9] method requires a high electromagnetic field strength; the bearing must be magnetized before detection can be performed. This may lead to a buildup of metal debris dislodged by friction during bearing operation, which accelerates bearing failure and reduces bearing life. Additional conditions that can constrain the data in the autoencoder such that they are only approximately optimized are required By introducing these constraints, the autoencoder can learn effective data feature representation

Objectives

Methods

Findings

Conclusion