The work of this dissertation is supported by National“863”High—tech R&D Proj ects of“the Key Technology in FMS--Stereo Vision System”．The sample robot system is designed for robot to work in Flexible Manufacture System．It works by vision and can recognize complex objects，position objects and track objects quickly in a large work space．The work of dissertation is an important part of the project，the main context iS as follows： Firstly，the general research study in the field of robotics is reviewed，and the object tracking and visual servo are studied in detail． Secondly，an method based on plant square for camera calibration is presented． Compute the relationship with the four points on square and the image proj ection points to obtain the intrinsic and extrinsic parameters of cameras．Then use the method to calibrate the robot hand—eye relationship． Thirdly,a method based on local HSV image and shape of obj ect to recognize object is proposed．After the color segment，use the knowledge of shape of objects to recognize balls，squares or hexagon parts． Fourthly．use the D—H model method to build the model of five—freedom robot to obtain the solution of the motor angle．According to the system，we built the tracking strategy to track objects．The correction is testify the tracking experiments． Fifthly，to control the sample robot system，a structure of a novel fuzzy PID controller is introduced and the nonlinear characters were analyzed．It uses a one-input fuzzy inference with three rules．The nonlinear gain properties of one—input fuzzy PID controller were analyzed by linear method that the nonlinear gain properties could be simulated by several lines，It can quantify and simplify the nonlinear gain．In addition，the stability is studied by the circle criteria．
修改评论