The Intelligent Grasp Of The Monocular Vision Servo Robot Based On Fuzzy Neural Network

As the science and technology growing, the demand of the robot's performance become higher, robots should not only be capable of implementing normal action, but also be capable of having human action, such as implement human visual, human thinking, human learning, human independent and so on. To satisfy such demand, this paper researches the model of the intelligent grasp of the monocular vision servo robot deeply, using the fuzzy neural network to identify the model of the intelligent grasp of the monocular vision servo robot.Firstly, the general research situation in the field of robotics and the development trend are reviewed. It also reviews the research content of the robot vision and the summary of the fuzzy neural network. Then the paper discusses the kinematics of the vision servo robot and the vision servo systems. Following the paper takes an in-depth study of the technology with the object identification and the fuzzy neural network. Finally this paper implements the intelligent grasp of the monocular vision servo robot.In the kinematics of the vision servo robot and the vision servo systems, this paper uses the method of D-H and the imaging geometric of establish the model of the positive and negative kinematics of the vision servo robot, it also analyses and compare the system of position based visual servo and the system of image based visual servo, the static visual servo system and the dynamic visual servo system.In the field of object identification, this paper uses the adaptive-middle filter to eliminate the effect of the salt and pepper noise to the image. At the same time, this paper uses the adaptive-weiner filter to eliminate the effect of the Gauss noise to the image. Then this paper uses the FCM to segment the object and the background and uses mathematical morphology to smooth the object. Finally through the analysis of the statistics, this paper obtains the eigenvector which can characterize the object.To model the model of intelligent grasp, this paper uses a six-layer fuzzy neural network and the three-stage hybrid learning algorithm to acquire a method of an automatic data-driven Visual-Degree function based on the fuzzy neural network.As to implement the intelligent grasp, this paper introduces the hardware of the monocular vision servo robot. The communication principle between the PMAC motion controller and the master and the communication principle between matlab and VC++ are studied. It uses VC++ to build the HMI to show the motion of the robot. Then the paper gives examples to implement the intelligent grasp of the monocular vision servo robot.