Vision guided ball-beam balancing system using fuzzy logic

Determining the behavior of a controller involving nonlinear systems is very difficult to test and develop. As a benchmark to test this controller, the ball-beam balancing problem has been used due to its inherent nonlinear dynamics. This paper presents a new ball-beam balancing system that uses camera vision to determine the position of the ball on top of the beam and its velocity. A fuzzy logic algorithm is used to control the motion of the beam to position the ball at its center. In this type of controller the mathematical equations describing the dynamics of the ball and the beam is not required. On line experiment depicting the real time behavior of the system is presented. Results show that the vision system and the proposed controller developed is accurate, adaptive, and robust.