Trajectory Optimization Of 7-DOF Humanoid Manipulator For High Speed Movement Target

The trajectory planning of humanoid manipulator must be carried out before setting out to do tasks.It is not high-demanding for humanoid manipulator to achieve a simple task.However,a task with high speed target like playing ping-pong or baseball makes the study of humanoid manipulator's trajectory-planning essential.The redundancy of the humanoid manipulator gives it more flexibility and larger work space,meanwhile,it brings huge challenge for the humanoid manipulator to realize fast motion.Focusing on the humanoid manipulator,this paper mainly studies trajectory optimization of 7-DOF humanoid manipulator for high speed movement target based on the kinematic of humanoid manipulator.Firstly,this paper introduces the hardware system and control system of 7-DOF humanoid manipulator.According to the operation requirement of 7-DOF humanoid manipulator,the motor of the single joint is taken for example to be configured.It provides good experimental condition for later theoretical research.Secondly,the 7-DOF humanoid manipulator joint structure is established and the forward kinematics model is analyzed.The solution to the inverse kinematic problem is reached by geometric analytical method.Considering the physical constraint of the humanoid manipulator for high speed movement target in operation,a "target operation level"is defined as optimization goal and traversing method is used to search the redundant variable within edge-restraint condition,then other joint angles of the humanoid manipulator are confirmed.The effectiveness of this method is verified by simulation experiments.The Jacobian matrix is computed and with which the angular velocity of each joint of the redundant humanoid manipulator is calculated.Thirdly,the optimal energy trajectory of the humanoid manipulator for the high speed movement target is choosen as the optimization goal and cubic polynomial interpolation is be used in the articulation space.Considering the physical constraint conditions of the humanoid manipulator in operation,then particle swarm optimization(PSO)is utilized to find the redundant variables of the trajectory parameterization to obtain the optimal energy trajectory.A 7-DOF humanoid robot for ping-pong playing is adopted as an example,and the method is employed to solve the trajectory planning problem of humanoid manipulator for ping-pong hitting.The trajectory before and after optimization is analyzed by numerical simulation and actual humanoid manipulator testing results,and the energy optimization effect and the satisfaction of constraints is verified.Lastly,researching on the process of operating rapidly for the high speed movement target of human being,two minimization performance criteria,i.e.energy consumption and configuration comfortable level of the humanoid manipulator are selected as the optimization goal.Considering the physical constraint conditions of the humanoid manipulator in operation,then multi-objective particle swarm optimization(MOPSO)is introduced to acquire the redundant decision variables of the trajectory parameterization to obtain the optimal Pareto set.As such,according to the decision requirements the decision maker can extract a compromised non-dominated solution that can be considered as the corresponding "knowledge" information of the multi-objective trajectory optimization.The 7-DOF humanoid manipulator for ping-pong hitting is adopted as an example,and this method is employed to solve its multi-objective trajectory planning problem.Numerical simulation and actual humanoid manipulator testing results show that this multi-objective trajectory planning method can approximate preferably the real Pareto front,and the selected non-dominated solution can be used as the "knowledge" data for on-line planning the humanoid manipulator trajectory that is similar to the motion characteristic of human arm,which can avoid the weight parameters setting in the traditional weighting method,and demonstrate the effectiveness of this method.