Research On Human-robot Control Method For Manipulator Based On Data-driven

With the development of global manufacturing process and technology,robot has gradually become one of the most important element in modern industrial.Robots are also the core force for enterprises to improve product quality and transform and upgrade to intelligence.Manipulator is the most common industrial robot in the manufacturing industry,which is widely used in cutting,assembly,stacking,spraying and other occasions,and is also widely involved in medical surgery,space exploration,military reconnaissance and other fields.With the arrival of the Fourth Industrial Revolution and the proposal of Made in China 2025,the global manufacturing production will usher in a huge change.Data-driven control and artificial intelligence will lead the future development,and the new generation of robotics technology has also become popular,This paper takes the 6-DOF industrial manipulator as the research object.Analyzing the kinematics and dynamics of the manipulator,proposing a data-driven trajectory tracking method of the manipulator,and then studying the human-robot control,which realizes the intelligent compliant control of the manipulator.Firstly,this paper summarized the framework of man-robot integrated hierarchical control structure for manipulator and analyzes the cartesian space trajectory planning method,then the mechanical arm collaboration within the working space of obstacle avoidance motion planning for research.This paper also proposes a adaptive step length of RRT algorithm,which improvements on the node extension,strengthen the guidance of planning algorithm and optimization of the planning speed.Finally,a smooth planning path is obtained by cubic B-spline curve fitting.Secondly,considering the strong coupling and non-linearity of freedom manipulator,we propose a data-driven trajectory tracking control method.Solving the dynamics equation of the 6-DOF manipulator,and using the Taylor method to obtain the discrete treatment of the dynamic equation,thus obtaining the state space model.At the same time,in order to improve the predictive controller effect,using the particle swarm optimization algorithm to solve the optimal problem in predictive control.At last we carry out a comparative test,which to verified the superiority of the data-driven control.Finally,studying the human-robot integration control method of the manipulator,which propose an active compliance control method based on man-robot integration,which combines the data-driven control method with the position-based impedance control method,so that the manipulator can simultaneously complete position tracking and force tracking.Then verifying the correction of the control method with MATLAB simulation platform.