Improved Forecasting Compensatory Control through Kalman Filtering

Abstract

The deformation of large thin-walled parts during the cutting processing will decrease the part accuracy, and on-line error forecasting and compensation control is usually used. The forecasting compensatory control (FCC) depending on modelling technique is usually helpful for some regular deformation. Random deformation of weak rigid thin-walled parts in the cutting process cannot be compensated easily. This paper develops an improved forecasting compensatory control method based on Kalman filtering algorithm to improve the prediction accuracy. The Kalman filtering algorithm produces the estimation of the real deformation based on the measured deformation data, and the statistical noise in measuring and cutting process modeling can be reduced. The effectiveness of the proposed method is validated with simulation examples.