58 / 2019-12-11 08:15:00

Remaining useful life prediction of bearings based on Continuous Hidden Semi-Markov Models

CHSMM; bearing; remaining useful life; prediction

IEEE SAM 2020: Full Regular Papers > Performance evaluations with experimental data

Forecasting methods based on models have been widely used in failure prediction of power equipments. The most representative one is the Hidden Markov Model (HMM), which applies probability theory. The Hidden Semi-Markov Model (HSMM) is an extended form of the HMM, and it has a higher accuracy in classification, which can be able to directly estimate remaining life by introducing state duration parameter to describe the degradation of a system. This paper gives a prediction method based on Continuous Hidden Semi-Markov Models (CHSMM), and confirms its feasibility in remaining useful life prediction of bearings from using the loading test data.