Kinematic control of a 5-DOF manipulator using adaptive neuro-fuzzy inference system
In this paper, a novel approach is introduced in order to control the position of an industrial S-DOF manipulator, due to severity of using algebraic and geometrical solutions in solving inverse kinematic problem (IKP). This method is based on training the robot at several points within its workspace. For this purpose, the end effector is moved to different positions by applying different joint angles to the motors and then sets of data consisting of points and joint angles resulting them, are gathered for each joint using Denavit-Hartenberg forward kinematics method. Then Anfis is used to train fuzzy logic controller for each joint. The controller is trained to transform Cartesian coordinates to joint space coordinates which is called inverse kinematic. So all joints of the manipulator will be independently controlled by a fuzzy controller. The mechanism is also modeled using SimMechanics in order to perform a real-time simulation and verify the system performance. Simulation results illustrate high capability of Anfis method in kinematic control of robot manipulators.
Manipulator inverse kinematic Anfis SimMechanics Denavit-Hartenberg
Pedram Safi Samghabadi Ablolfath Nikranjbar
Islamic Azad University - Karaj Branch Faculty of Engineering Karaj, Iran
国际会议
2010 International Conference on Measurement and Control Engineering(2010年IEEE测量与控制工程国际会议 ICMCE2010)
成都
英文
351-355
2010-11-16（万方平台首次上网日期，不代表论文的发表时间）