Predictive Contour Control With Adaptive Feed Rate

Contour control is an important issue in motion system development. In this paper, a contour-control methodology combining predictive control and adaptive feed rate is proposed. To reduce the computational time required for online optimization, a simple unconstrained model predictive controller is first designed to perform biaxial contour-position control. The performance index introduced in this paper allows the designer to manipulate the importance between contour error, axial tracking error, and control usage. The controller is then implemented to track diamond and free-form contours. The results demonstrate that the controller is capable of not only tracking both contours with high precision (contour and axial) for steady motions, but also improving contour precision during transient periods. To further reduce the contour error during transient periods, an adaptive feed-rate scheme is proposed, which utilizes the predicted contour error to adjust the feed rate online. This adaptive feed-rate scheme is experimentally validated using diamond and limacon contours. The results demonstrate that in comparison with constant feed-rate schemes, the proposed adaptive feed-rate scheme is capable of significantly reducing the transient contour error at high feed rates, while maintaining comparable tracking performance at low feed rates.