A Model Predictive Contouring Error Precompensation Method

Abstract

Contouring error has a significant effect on machining accuracy. In this paper, a model predictive contouring error precompensation method is presented that performs model-based error precompensation with feedback and constraints. The proposed method uses the predicted contouring error without compensation obtained by simulation, as the reference trajectory of the model predictive compensator. It then calculates the compensation signal using a constrained model predictive control algorithm and adds it to the original reference position as the new system input. Experiments were conducted on an X-Y table to compare the proposed method with a traditional cascaded proportional-integral (PI) controller and two existing contouring error precompensation methods. The experimental results show that higher contouring accuracy can be achieved with the proposed method for both sharp corners and free-form curves.