Research On Dual Arm Coordination Control Method Of Humanoid Robot For Flitting Task

Humanoid dual arm robot can imitate human hands to complete the operation tasks such as carrying,screwing valves and so on.With its distinct operational advantages and robustness and flexibility in the operation process,it has been widely used in service robots,special robots and other fields.It is still recognized as a major challenge in the field of control to use both arms for stable operation,such as clamping or handling objects.Based on the humanoid dual arm robot platform developed by the scientific research project of XXX operation system,this paper studies the dual arm coordinated control method for load handling task.Firstly,a software control system is built based on ROS robot operation platform,which is divided into user layer,middle layer and robot layer.The middle layer includes: fast partition layer,module function layer and hardware communication layer.This control framework is not only suitable for dual arm coordinated handling tasks,but also can be easily extended to the development of navigation,obstacle avoidance,voice interaction and other functional modules.Secondly,the kinematics model of the two arms is established.For redundant degree of freedom manipulator,efficient and reliable inverse kinematics solution is the focus of kinematics analysis.In this paper,the arm angle method is used to calculate the inverse solution of the manipulator position level,and the gradient projection method is used to obtain the inverse solution of the manipulator velocity level.Based on the quintic polynomial interpolation and the gradient vector in the process of solving the inverse solution of velocity level,the objective function is optimized to realize the collision free Cartesian space trajectory planning.Then,it focuses on the two arm coordinated control method for handling.In the task of double arm tightly coordinated operation,the dynamic model of double arm and object is established as a whole.Through grasping matrix,the desired resultant force exerted by the manipulator to the object is dynamically distributed to the two manipulators.Using the impedance control strategy,the manipulator and the contact part of the object are equivalent to a spring mass damping system.In the control framework,the impedance of the manipulator is used as the inner control loop,and the environmental impedance in contact with the manipulator is used as the outer control loop.Finally,a simulation model is built in the Adams simulation environment,and the above algorithm is used to simulate the dual manipulator coordinated handling task,which verifies the feasibility of the control theory and the safety of the control framework application.Experiments are carried out on the real robot platform.In order to ensure the smooth implementation of the two arm operation experiment,firstly,the position control accuracy of the manipulator is tested,and the handling experiment process is designed.Finally,the experiment of two arm coordinated handling of rigid objects is successfully completed.In the experiment,the two manipulators can successfully clamp the object to be operated,and maintain the set target contact force to transport the object to the target position smoothly,which verifies the effectiveness of the control algorithm...