A genetic algorithm-based parameter-tuning algorithm for multi-dimensional motion control of a computer numerical control machine tool

Abstract

This paper addresses an automatic parameter-tuning algorithm for the multi-axis motion control of a computer numerical control (CNC) machine centre. The traditional approach to tune the control parameters in the multi-axis machines is to tune each axis independently. Some high-end-precision machines offer cross-axis motion parameters for impedance compensation but this is usually not satisfactory for practical purpose. Because each axis on the machine centre contributes to more than one working plane, obtaining the optimal performance for motions involving more than one plane often results in axis coupling. This paper introduces a systematic method to tune the servo parameters for multi-axis motion control. The tuning algorithm is based upon an intelligent genetic algorithm (GA) and the parameters are tuned for each work plane. The method optimized the multi-axis motion performance. A modified GA is also proposed to solve the convergence problem induced by a large number of parameters in multi-axis motion tuning.