Discrete-time model predictive contouring control for biaxial feed drive systems and experimental verification

In this paper we present a novel algorithm to model predictive contouring control for biaxial feed drive systems. model predictive control (MPC) refers to a class of model-based controllers that uses an explicit process model and tracking error dynamics to predict the future behavior of a plant, making it effective for machine tool feed drive systems that must achieve high-precision motion and are severely affected by friction, cutting force and changes in the workpiece mass. To improve contouring performance, we propose a new performance index in which error components orthogonal to the desired contour curve are given more importance than tracking errors with respect to each feed drive axis. Controller parameters are calculated in real time by solving an optimization problem. The parameters depend on the instantaneous slope of the reference trajectory and thus vary with time for curved reference trajectories, resulting in a time-varying controller. Weighting factors for the error components in orthogonal and tangential directions are used to adjust the error importance in each direction. In addition, to consider the required feed drive energy, the control inputs in both directions are included in the performance index. The effectiveness of the proposed control approach is demonstrated with an experimental biaxial feed drive system for circular and non-circular trajectories. The proposed contouring controller allows the feed drive to follow smooth curves and reduces contouring error.