Mult–Objective Performance Optimization For The Similar-SCARA Parallel Robot With Limbs Of Embedding Structures

With the development of science and technology as well as the improvement of industry,the requirement for the performance of industrial robots is becoming increasingly strict,so the multi-objective performance optimization of the robot has a very important significance.In order to improve the performance of the similar-SCARA parallel robot with limbs of embedding structures which is applied to wafer handing,this thesis presents a study of the performance analysis and the multi-objective performance optimization for this robot.Firstly,this paper introduces the research background of the multi-objective optimization of the similar-SCARA parallel robot,and outlines current research state of the parallel robot,performance analysis of the parallel robot and the multi-objective optimization algorithm.Then the significance of the multi-objective performance optimization for the similar-SCARA parallel robot is explained and the main contents of this paper is introduced.Secondly,the kinematics and dynamics of the similar-SCARA parallel robot with limbs of embedding structures is analyzed.The robot Jacobian matrix is solved,the Lagrange method is applied to get the dynamics equation of the robot,and according to the influence of the geometric error of the moving platform on the coordinate vector of the moving platform,the geometric error model is established.According to the related theory analysis,the four performance evaluation indexes of the robot including the work space,global condition index,natural frequency and error influence coefficient are got.Finally,based on the analysis of the performance of the similar-SCARA parallel robot,the multi-objective optimization of the four performance index is carried out.The optimization objectives,optimization variables and constraints are determined,and the multiobjective optimization model is established.A multi-objective genetic algorithm based on BP neural network is used to solve this problem,meanwhile,the structure of the BP neural network is optimized by genetic algorithm to improve the prediction accuracy and reduce the computation time.The MATLAB optimization function is written,the optimization parameters is set,and the multi-objective optimization model is solved.The validity of the optimization is proved by comparing the performance index before and after optimization.The ADAMS kinematics simulation is used to prove intuitively that the optimized performance of the robot is improved,by comparing the of simulation atlas before and after optimization.