Research On Motion Control Of Quadruped Mobile Robot With Dual-arm

With the continuous improvement of the current industrial production level,mobile robots equipped with arms are playing an increasingly important role in the field of manufacturing,and they have excellent performance in assisting humans to carry heavy objects and perform operations.Among them,quadruped robots equipped with dual-arm not only have good ability to traverse rough terrain,but the dual-arm system also facilitates the robot to operate on objects with irregular shapes,high mass and large volume.However,in the process of manipulation of quadruped robot with dual arms,the continuity and coordination of its movement is important prerequisites to ensure the robot completes its tasks due to the complexity of the operating environment.Based on the above background,this dissertation realizes the path planning of the grasped object and the torso in the task space with a sampling-based planning method,and obtains the path in the joint space according to the robot kinematics analysis to realize the motion planning.The singularity of the closed-chain system is analyzed and handled during the planning process.Finally,a controller based on inverse kinematics is designed to achieve joint control of the robot in a simulation environment.The specific research work is as follows:(1)Aiming at the task space analysis and simulation construction of the robot,we establish its basic coordinate systems according to the physical parameters.The task space of the quadruped robot with two arms is divided into the closed-chain system task space and the quadruped robot task space.The closed-chain system task space is treated as the overall task space for robot motion planning.The robot and the environment are modeled in the Webots simulation software,and a co-simulation environment of Webots and MATLAB is built.(2)Regarding the kinematic analysis of the quadruped robot with dual-arm,we establish kinematic models for the legs,arms,and torso,and designs a method for the inverse kinematics of the 6-DOF manipulator based on Mayfly Algorithm(MA).Firstly,the kinematic equations for the legs and arms are established by the Denavit-Hartenberg(D-H)method.The inverse kinematics of the legs are solved using the algebraic method,and the forward and inverse kinematics of the torso are derived based on the basic coordinate system.When using the MA to solve the inverse kinematics of the 6-DOF manipulator,the error between the actual and expected end-effector poses is used as the fitness function for the MA.The accuracy and efficiency of the inverse kinematics have been improved through the iterative solution by algorithm.Finally,the method for the inverse kinematics of the 6-DOF manipulator based on MA is experimentally validated in the simulation environment.(3)In response to the singular point of closed-chain systems,we analyze and construct the Jacobian of the closed-chain system to determine the closed-chain singularity.The Jacobian pseudo-inverse method based on Damped Least Squares(DLS)is improved by changing the adaptive damping factor and optimizing the optimal damping factor to improve the algorithm’s solution efficiency.Designing a closed-chain singular point avoidance algorithm based on DLS to solve the Jacobian pseudo-inverse,and further improving the solution accuracy through error feedback.And it is verified experimentally in the simulation environment.(4)Regarding the motion planning and control of the quadrupedal and dual-arm robot,a collaborative planning method for the dual-arm system and the quadruped robot is designed,and experimental verification is achieved through joint control in the simulation environment.Based on the task space analysis of the quadruped and dual-arm robot,the path of the grasped object is planned in the overall task space using the Rapidly-exploring Random Trees(RRT)algorithm,and the feasible workspace of the torso is analyzed according to the kinematic constraints.A sampling strategy is designed to plan the torso motion path,achieving collaborative planning of the quadruped bimanual robot.Finally,a controller is designed based on the joint angle information obtained from the joint position sensors,and the robot’s motion control is experimentally verified in the simulation environment.