Integrated prognosis of AC servo motor driven linear actuator using Hidden Semi-Markov models

Abstract

Motor prognosis has great benefit for motion control systems. It is desirable to develop the so-called integrated prognosis capability for motion control products, i.e. with only the existing sensors of a motor drive. This paper presents a scheme of integrated wear prognosis for linear actuators driven by AC servo motor with only current measurement. The Hidden Semi-Markov Model (HSMM) was adopted as the prognostic model. Compared to a single observation for a state in the hidden Markov model (HMM), a state in an HSMM generates a segment of observations. Therefore, HSMM structure has a temporal component. For the HSMM based prognostic model, the states correspond to different levels of health status. The duration of a health state is modeled by an explicit Gaussian probability function. To avoid the underflow of computation, the modified forward-backward training algorithm was used to estimate the model parameters. The expectation-modification method was used to estimate the state transition probability matrix, initial state distribution, and observation probability matrix in the model. The trained HSMMs can be used to predict the remaining-useful-life (RUL) on the actuator. The proposed method is demonstrated through the simulation of the wear process of an AC motor driven linear actuator.