Research On Motion Planning Method Of Wearable Exoskeleton Arm Based On Dynamics

Exoskeleton robot system is a combination of sensing,control,information coupling,mobile computing and other technologies.It can effectively enhance human body functions.The effect of assistance and rehabilitation will be affected by the integration of human-machine and scientific of mechanical structure.In the research process of exoskeleton robots,control strategy is one of the core technologies.Therefore,it is the main direction of the exoskeleton robot technology in the future to develop more efficient control strategies.It can improve the practicality of exoskeleton robots.In order to ensure the safety and flexibility of exoskeleton arm,dynamics-based control and motion planning methods are emphatically studied in this paper.Based on the comparative analysis of the domestic and foreign research status of related control technologies,the force-free controller is designed to achieve the servo mode of the exoskeleton arm when there is no task.According to the assist mode of the exoskeleton arm,an impedance control strategy is developed.And the trajectory planning method based on dynamic motion primitives is improved.While the assistance from the exoskeleton arm is provided,the regular reciprocating movement between the dynamic targets is realized.The main contents of this article are as follows:(1)The kinematics of the exoskeleton arm is analyzed by the DH model.And the effectiveness is verified by the principles of plane geometry.Three methods are used to solve the Jacobian matrix of the exoskeleton relative to the base coordinate system.A method for solving the singular configuration using the relationship between the rank of the matrix R and the(R+1)order sub-determinant is proposed.The correctness of the singular configuration is verified by the condition number,minimum singular value,isotropy and other dexterity indexes,and the advantages and disadvantages of each dexterity index are analyzed.(2)Using the law of conservation of energy and the equivalent principle of flexibility,a dynamic model of the exoskeleton arm is constructed.The mass matrix,gravity matrix and dynamic equation are solved respectively.The dynamic model of exoskeleton arm was constructed.In order to better realize motion control and motion planning based on dynamics,current detection is used to estimate torque instead of sensor.Parameters such as motor sensitivity,static friction,viscous friction,and coulomb force are identified.The current for gravity compensation is derived.The relationship between DC motor current and torque is a prerequisite.(3)The zero-force controller is designed and its rationality is analyzed.The impedance control strategy is formulated.And The influence of parameters on system stability is analyzed.Then the simulation to track the compensation torque is made.For the dynamic motion primitive trajectory learning method,the optimization of extending the basis function center and changing the basis function generation method are proposed.Then the efficiency and accuracy of trajectory learning can be improved by the above methods.The corresponding experiments and analysis were carried out on the independently developed wearable upper limb exoskeleton robot experimental platform.