Design of a sub-optimal digital controller for an antenna control system

This thesis deals with the design of a digital controller for a relatively complicated antenna system which contains several nonlinearities and is of a degree of complexity which makes a mathematical analysis of the control problem extremely difficult if not impossible.;A simplified model which gave satisfactory outputs for a variety of step inputs was obtained. The state equations were derived from this model and used in attempts to determine the theoretical optimum control for a step input command to the system. Two methods were used but both failed to yield useful results. Thus, instead of employing the theoretical solution as a standard of performance, a graphical approach was taken.;It was found that the control system was velocity limited for large step inputs. The graph of the position error of the system with a bang-bang controller thus represented an easily understood and relatively precise measure of the degree by which other control schemes fell short of achieving a true optimal control.;The type of control algorithm which was developed is easy to implement and has an almost minimum time response with no overshoot for step input commands. Antenna velocity is normally zero at both endpoints. In addition the algorithm makes possible the elimination of the compensation network. This is presently used to condition the command signals which are passed from the computer to the control system.;The problem was approached by first simulating the system on a hybrid computer. Then, before the implementation of the experimental controller, attempts were made to obtain a representative mathematical model which could be used as a yard stick to measure the digital controller performance against a theoretical optimum.