Research On Structure Optimization Of SCARA Robot And Multi Domain Simulation

SCARA robots are a kind of lightweight industrial robots. Since they have suchfeatures as fast moving, accurate positioning, they have been widely used in assemblyfield In recent years, China’s prosperity in electronic industry significantly promotes theassembly industry. Under this circumstance, SCARA robots begin to carry on moreassembly operations. In this way, there are both high academic and economic value inanalyzing and optimizing SCARA robots.In this thesis, a SCARA robot which was developed by Haier Laboratory of theState Key Lab of Robotics and System was optimized and simulated with multi-domainsimulation approach. First of all, the kinematics and dynamics features of SCARA havebeen carefully analyzed. The Jacobian matrix and Hessian matrix were deduced. Basedon Lagrange method, rigid-flexible interaction equation was established. Secondly,determine the optimal primary goal is to improve the dynamic performance of the robot.Through the analysis of robot mechanism, performance SCARA robot dynamics,kinematics performance indicators and static stiffness performance indicators werededuced. The dynamic performance indicator was served as the first optimization goal.The kinematic performance objective function was served as the second goal. The staticstiffness performance indicators and work space were served as constraints. Theparameter of large arm and forearm were served as the optimization variables. On thebasis of the above setting, a robot institutional dimension optimization model wasestablished. The NSGA-II method was used for solving multi-objective optimizationproblem. In the Pareto solutions, a solution which can dramatically improve theperformance of the robot dynamics and can’t significantly impair the selected kinematicperformance was chose. The selected optimal solution was substituted into robotdynamics and kinematics equation to calculate theoretically. The result proved that therobot dynamics optimized performance was greatly improved, the kinematicsperformance was slightly improved and workspace basically unchangedBased multi-domain simulation modeling specification Modelica, a SCARA robotjoint multi-domain simulation model which including institutions, joint drive systemcomponents, trajectory planning system components was established in Mworks. Themodel was used to optimize the results of verification. Multi-domain co-simulationtechnology can solve complex system simulation in various fields generated when thecoupling problem. This method can create a complex model including mechanical,electrical, control and other subsystems. This modeling approach is very suitable for therobotic system modeling.Finally, I conducted experiments with the SCARA robot to verify the result of theoptimization experiment.. Under the same conditions the angle of each joint, the maximum speed and maximum acceleration were assigned. I collect the output of themotor output shaft. This experiment verified the accuracy of the multi domainsimulation of SCARA. Later, given the same experimental conditions, by comparingbefore and after optimization of joint torques confirmed through the robot’s body sizeoptimized dynamic performance has been improved.