Implementing ball balancing beam using digital image processing and fuzzy logic

Fuzzy logic is a paradigm for an alternative design technique, which can be applied in both linear and non linear systems for embedded control. It is a problem-solving control system methodology that can be built into anything from small, embedded controller and hand-held products to large computerized process control systems. This paper proposes and analyses methods to control a ball balancing system using fuzzy logic. It is a nonlinear dynamic system which is quite difficult to control using conventional methods. It has delayed feedback associated with control actions; however, control system based on fuzzy logic requires less mathematical derivation in design and has a high noise tolerance. Image of the ball taken from an audio/video (AV) camera is processed to determine the coordinates of its centre. This input data is fed into the 68HCS12 microcontroller, which has the most efficient implementation of fuzzy logic instruction sets. Built-in PWM channels, A to D converters, 32 kb flash memory and in-system programming make this controller ideal for the system considered. The controller processes this input using the fuzzy logic rule base and sends consequent signals to the servo motor controlling the beam motion