Trajectory control of analog servo motor with limited state information using estimated discrete time model

Abstract

An estimated discrete time model from analog servo motor with limited state information was obtained. The process is divided in two main parts. First, we build diagram blocks consisting DC motor and its position controller. Next, we find the whole transfer function derived from the diagram blocks. Second, we obtained sampled-data from experiment to estimate the coefficients of simplified discrete time transfer function. In doing this, we use Gauss-Newton algorithm. We proposed a control method to control motor trajectory for the obtained model without directly controlling its velocity and only using position as a reference. Simulations and experiments show that our method achieve lower acceleration/deceleration at initial/final time, faster settling time, and less kinetic energy than its built-in controller alone.