Research On Motion Planning And Control Of Redundant Dual-arm Robots

With the development of Electronic industry,collaborative robots which is lightweight and flexible have received more and more research.As a kind of cooperative robot,dual-arm robot has achieved much progress in recent years.Compared to traditional single-arm robots,Coordinated operation of two arms help improve operation accuracy.Both arms have a redundant degree of freedom,through this feature,a better path can be obtained by choosing some optimization performance index.With that,the robot can achieve obstacle avoidance and singularity and so on.This paper mainly studies the theory of a dual-arm robot with 7 degrees of freedom,the work can be divided into three parts: kinematics,motion path planning and motion control.First,kinematics is studied which was modeled by a single arm.Forward kinematics was solved by D-H method.As for inverse kinematics,FABRIK algorithms and pseudo-inverse jacobian were used to solve the inverse solution.It is discussed about advantages and disadvantages of these two methods.Simulation was implemented in MATLAB to verify algorithm validity.The relationship between two-armed coordination of movement constraints is studied.Simscape Multibody software was used for dynamic simulation.Then,the performance index of the motion planning are introduced,and the collision detection methods of the robots are analyzed.The motion planning methods in the high-dimensional space are studied,the rapidly exploring random tree method and the probabilistic roadmap method are used.And through simulation result the run time in algorithms is compared.Through MoveIt! Software,3D simulation with obstacle avoiding is documented.Finally,in the motion control section,many control methods are compared about advantages and disadvantages.A simple 7-degree-of-freedom dynamics model is built using the dynamic algorithm library.and feedback control is performed through Simulink.The design of fuzzy controller was carried out in MATLAB,and fuzzy control simulationwas performed on one direction on the end joint.