Optimum predictive and feedforward control of a precision linear stage using genetic algorithm

In the area of factory automation, vision based leads inspection systems are increasingly used to ensure that the lead forms of the integrated circuits meet the PCB assembly requirements. Typically, it comprises a camera system and a precision linear stage. In this paper, the development of a motion control system used in the authors' prototype leads inspection system is presented. High performance is obtained by using a permanent magnet DC linear motor as the actuator for the linear stage, which eliminates the needs of mechanical transmission from the rotary to linear motion. Fast and accurate servo response is achieved by using a state space predictive controller with feedforward compensation of friction. The controller is optimally tuned using the genetic algorithm. The experimental system has a positioning accuracy of 1 /spl mu/m and a peak speed of more than 1 m/s. Some experimental results are presented.