MODEL-FREE METHOD FOR TIME VARYING DYNAMIC MEASUREMENTS IN CONTROL SYSTEM

Abstract

In this paper, a model-free method for time-varying dynamic measurements in a control system is presented. As an example, the dynamic mass-measurement process is examined. The method is based on the on-line estimation of time-varying parameters of autoregressive model by a recursive least square method with a constant trace of the covariance matrix. The model order selection is performed by Akaike’s information criteria. The performance of the method with respect to the variance of measurement noise is empirically tested by simulation experiments. For the aim of comparison, the Kalman filter for estimation of unknown measurement is designed. The simulation results show the advantage of the model-free method.