Predictive control and intermittent setpoints are proposed to overcome the dead time that problem occurs in a new class of high precision position sensor for manufacturing equipment. In place of a rotary encoder or linear glass scale, a combination of a digital camera and a Liquid Crystal Display (LCD) screen is used to actively monitor two dimensional position changes on an XY table. In order to achieve precise spatial resolution, an actively-controlled planar pixel matrix is used as the tracking target for the system. A digital camera senses the location of the moving image displayed on the LCD screen and provides 2 dimensional position feedback. Thus, the timing and the quality of the visual feedback to the controller are the significant factors to determine the accuracy of the system. Due to the long image processing time, the vision feedback of the actual position of the stage is delayed. At the same time, with the slow frame capturing rates of the camera, dead time occurs between consecutive acquisitions of feedback signals from the vision system to the motion controller, which is detrimental to the performance of the system. Hence, studies and detailed analysis on different dead time compensation strategies and path planning algorithms have been performed to select the optimal strategy to address these challenges. Based on simulation results, a proposed method for integrating predictive control with virtual intermittent setpoints algorithm to mitigate dead time problem is presented in the final section of the paper.
