Comparative Study of Data-Driven Models in Motor RUL Estimation

Abstract

The extremely complex loading conditions on the clutch of a four-wheeled passenger vehicle frequently results in malfunction of the motor. The latest diagnostic methods for detecting the initiation of device failure have proven to be unreliable. The present research has been carried out to demonstrate the state of health of the motor on the basis of a nonlinear real time estimation approach. In order to fulfil this task, a systematic review was undertaken of the unscented particle filter (UPF) approach to handle the evolved noisy signal with in real time. Research facilitates the modeling of nonlinear behavior of elements via state-space equations embedded with a set of available real time measurements. The remaining useful life (RUL) of the motor (system) as a distribution function is estimated. The study highlights that the state space framework provides better results than the degradation modeling scheme to forecast the RUL of the system.