Design And Development On Ball&Plate Control System Based On Computer Vision

"Design and Development on Ball&Plate Control System Based on Computer Vision" is part of the 985 program in Control Theory & System Laboratory. The goal is to build Ball&Plate control system based on computer vision for the course of control theory for undergraduate。The thesis introduce the process of designing and manufacturing the hardware and developing software of interface of control system.. We design a fuzzy controller of Ball&Plate control system and simulate in matlab environment. Then we apply the result in the real apparatus. Ball&Plate control system is a complicated nonlinear and astable system. The exploitation of real Ball&Plate control system is blank in China. My thesis is introduce the whole process on how to manufacture the platform. It's mainly divided in two parts: hardware manufacturing and software development. Hardware designing include the shell , switch, laying power supply and CCD camera, choosing capture grab card and step motor control card, and connection between model and computer; software development include image processing in real time,step motor control program and friendly windows interface between user and computer. The process is as below: ball is rolling on the plate freely, and it's position is captured by CCD camera; Then the information of ball will be saved in memory; program will read the image data from memory and analyze relative position and speed of ball; then acquire output by fuzzification, fuzzy inference and defuzzification, output that include frequence,quantity and direction of pulse will be send to step motor drivers via PCL839 card. Plate rotate by control signal to control ball position as planned by user.Due to nonlinear ,we choose the fuzzy control method for Ball&Plate control system. Based on Ball&Plate control system, we creatively design a fuzzy equivalent of PD controller. The response of ball behavior is satisfied under matlab/simulink. Based on the simulation in matlab , we use the result in the real world and improve it's performance, and achieving stabilization control.