Application of grey predictor and fuzzy speed regulator in controlling a retrofitted machining table

Abstract

In this work, the control plant is an a.c. servo motor retrofitted traditional manually operated milling machine with a lead-screw transmission system. This obsolescent milling machine has non-linear time-varying behavior due to obvious backlash and irregular coulomb friction of the sliding surfaces. The system model is difficult to derive and identify for classical control design. In order to achieve the objectives of reducing the tracking error and path error and increasing the motion speed, a combination of PI control with grey prediction, cross-coupling and feedforward control loops and fuzzy speed regulator is proposed to control this machining table. The grey prediction loop is employed to improve the robustness and tracking accuracy of the control system. The fuzzy speed regulator is used to achieve the maximum machining speed under a specified error tolerance. The experimental results show that this control method achieves satisfactory performance of transient response, tracking and robustness under the influence of about 0.4 mm backlash on each axis and large stick-slip friction. This performance verifies the applicability of this economical approach to the automation of traditional milling machines.