Using Genetic Algorithms to Optimize Control of a Ball-and-Beam System

Abstract

The purpose of this paper is to develop a methodology for using a Genetic Algorithm (GA) to tune a PID controller, which will stabilize a ball-and-beam system. A brief overview of GAs will be given followed by a short introduction of the ball-and-beam system, to which a GA will be applied. Next, the method of applying a GA to a PID controller for optimization is discussed. A conventional PID controller and an LQR controller will be designed for the purpose of evaluating the cost associated with these controllers against the cost associated with the GA optimized PID controller. The final results show that a PID controller tuned using the GA is more cost efficient than a conventionally tuned PID controller, but less cost efficient than a conventionally tuned LQR controller.