Research on Electrostatic Monitoring of Tribo-Contacts with Dynamic Adaptive Fusion Method

Ruochen Liu,Shuangshuang Gu,Jianzhong Sun,Jinwu Zhang,Han Wang,Ricardo Escobar

doi:10.1155/2022/8952349

Abstract

Electrostatic monitoring is a unique and rapid developing technique applied in the prognostics and health management of the tribological system based on electrostatic charging and sensing phenomenon. It has considerable advantages in condition monitoring of tribo-contacts with high sensitivity and resolution. Unfortunately, the monitoring result can be affected due to the switch of operating conditions that reduces its accuracy. This paper presents a dynamic adaptive fusion approach, moving window local outlier factor based on electrostatic features to overcome the influence. Life cycle experiments of rolling bearings and railcar gearbox were carried out on an electrostatic monitoring platform. The MWLOF method was used to extract and analyze the experimental data, combined with the Pauta criterion to judge wear faults quantitatively, and compare with other feature extraction results. It is verified that the proposed method can overcome the influence of changes in working conditions on the monitoring results, improve the monitoring sensitivity, and provide an accurate reference for friction and wear faults.

Highlights

Online condition monitoring is a significant concern in improvements of prognostics and health management (PHM) for mechanical equipment [1,2,3]
In the experiment of rolling bearing with the test rig introduced in Section 2 (Figure 2(a)), a life cycle test with two operating stages is conducted with electrostatic monitoring method
It is difficult to integrate various features from multiple electrostatic sensors and overcome the influence of operating condition changes for the system. is paper considered the characteristics of electrostatic signals and introduced a modified moving window local outlier factor (MWLOF) algorithm with mode change as a dynamic adaptive fusion method to indicate the performance and verified with experiments of rolling bearings and gearbox

Summary

Introduction

Online condition monitoring is a significant concern in improvements of prognostics and health management (PHM) for mechanical equipment [1,2,3]. After a series of experimental works, the electrostatic monitoring technique has been proved to be effective with its high sensitivity in condition monitoring of tribo-contacts [21,22,23,24]. Unlike conventional monitoring methods such as temperature and vibration that measure the secondary effects of wear, electrostatic monitoring is a direct measurement with detected electrostatic charges to the wear components by using electrostatic sensors [25,26,27]. It can provide the detection of severe failures and early warnings and reflect the degradation of mechanical systems

Methods

Results

Conclusion