Research on Motion Control of a7-DOF Manipulator
|Course||Detection Technology and Automation|
|Keywords||7-DOF manipulator obstacle avoidance planning fuzzy compensation computed torque control|
The redundant robot is kinematically redundant, so the joints of themanipulator have self motions when the position and orientation of the end-effectorare definite. Hence the redundant robot can get good dexterity and versatility.Therefore the redundant robot is significant to realize the complex operation. Thispaper took a kind of7-DOF manipulator as the research object, analyzed kinematics,dynamics, obstacle avoidance planning and track following control of theredundancy manipulator.The Cartesian coordinates of this7-DOF manipulator was established byDenavit-Hartenberg method. The forward kinematics was modeled throughhomogeneous transformation. Simultaneously, the algebraic-geometric method andthe gradient projection method were used to set up the inverse kinematics model onboth position level and velocity level respectively. The forward and the inversekinematics model were validated by simulating in the MATLAB.In order to realize obstacle avoidance, the obstacle model was simplified to theellipsoid, and the obstacle avoidance index was established based on thepseudo-distance, then the obstacle avoidance planning was established by gradientprojection method and constrained optimization method separately. Through thesimulation analysis, both methods could avoid obstacles effectively. However theobstacle avoidance index of the obstacle avoidance algorithm based on constrainedoptimization method was better. And the obstacle avoidance algorithm based ongradient projection method had higher accuracy, faster operation speed and moresmooth joint movement.The dynamic model of this7-DOF manipulator was established using theLagrange method, in order to obtain the relationship between motion and torque(force) on each joint. The virtual prototype of the manipulator was constructed byutilizing Pro/E and ADAMS. The dynamics simulations were carried out respectivelyin MATLAB and ADAMS, and the results of them were basically accord. Thesecould provide the dynamic model and the controlled plant based on virtual prototype for the track following control of the manipulator.The compute torque control with fuzzy logic compensation was presented bycompensating the uncertain part of the dynamic model via adaptive fuzzy logicsystems, which was to solve the problem that the traditional computed torque controlwas undesirable robustness when the dynamical model of7-DOF manipulator wasuncertain. ADAMS and MATLAB software was used to establish a co-simulationplatform and to simulate the manipulator and control system. The results showed thatthe control method could not only improve control accuracy, but also had greatrobustness.