Model predictive control to mitigate chatters in milling processes with input constraints

Abstract Due to the rigidity-flexibility coupling and self-excitation mechanism, chatters are often encountered in machining processes. They severely limit the productive capacity of machine tools, and lead to inferior work piece quality, cutting disturbances and quick tool wear. In recent years, the increasing industrial demand of high quality and high efficiency machining motivates us to develop a niche active control method to mitigate the chatter dynamics with input constraints. In this work, an active model predictive control (MPC) method for the milling process is developed such that the chatter-free domain of stable operation is substantially enlarged and a higher efficiency can be thus achieved. Therein, the complex perturbation dynamics including time-delay and periodical excitation is transformed into a linear time-varying (LTI) system, and afterwards both model-based prediction and receding horizon optimization are implemented by the proposed MPC scheme to address the system uncertainties and input constraints to guarantee the chatter-free stability and feasibility. Effectiveness and superiority of the proposed MPC are finally demonstrated by means of illustrative examples.