A Robust Model-Based Approach for Bearing Remaining Useful Life Prognosis in Wind Turbines

Wei Teng,Yankang Hu,Xin Cheng,Lei Song,Chen Han,Yibing Liu

doi:10.1109/access.2020.2978301

Abstract

Accurate remaining useful life prognosis of bearings in wind turbines can effectively help to schedule maintenance strategy and reduce operational costs at wind farms. Unscented particle filter is good at state tracking in nonlinear problem. A robust model-based approach based on improved unscented particle filter is presented to deal with bearing life prognosis in wind turbines, which involves: (1) The mean of sigma points after unscented Kalman transform is regarded as the particles in particle filter to guarantee the particles aggregation; (2) Several past measurements are utilized to estimate the likelihood function of current step; (3) Uniform distribution is adopted for resampling particles to make them diversity. The presented remaining useful life prognosis approach depends more on the measurement, rather than the initial parameters of degradation model, which makes it practicable for the on-site wind turbines. Three life-cycle bearings from wind turbine high-speed shafts demonstrate the effectiveness of the proposed approach.

Highlights

Wind energy has rapidly developed in the past decade worldwide
We focus on the bearings in wind turbine with high rotational speeds since abundant on-site cases exhibit that their degradation processes obey the exponential model as Eq (19)
Bearings remaining useful life prognosis plays a significant role in the operation and maintenance of wind turbines

Summary

INTRODUCTION

Wind energy has rapidly developed in the past decade worldwide. In China, by the end of 2018, over 221.6 GW of installed capacity of wind energy has been put into operation [1], contributing about 6% of the total power supply. W. Teng et al.: Robust Model-Based Approach for Bearing RUL Prognosis in Wind Turbines abundant historical life-cycle data that is hardly acquired in industrial applications. Unscented Kalman transform enables the nonlinear transfer of the mean and variance of particles, which generates proposal distributions that match the true posterior closely With these advantages, Acuña and Orchard [17] proposed a particle filter based failure prognosis via sigma points application to lithium-Ion battery state-ofcharge monitoring. Zheng and Fang [18] integrated unscented Kalman filter and relevance vector regression for the lithiumion battery remaining useful life and short-term capacity prediction Another significant work in RUL prognosis is the construction of health indicator. A robust model-based approach on the basis of improved unscented particle filter is presented to predict the remaining useful life of bearings in wind turbines. The posterior density p(xk |z1:k ) is estimated as

UNSCENTED PARTICLE FILTER

BEARINGS RUL PROGNOSIS BASED ON IMPROVED UPF

THE RESULTS OF RUL PROGNOSIS

Findings

CONCLUSION