Dual-arm Collaboration Technology For Home Service Robots

In recent years,the number of elderly people in China's population has gradually increased and the development prospects of home service robots have become broader.The home service robot has a humanoid dual-arm system.Compared with a single-arm robot,it has better collaboration capabilities and a wider range of service application scenarios.Dual-arm coordinated operation can effectively improve the operability,carrying capacity and reliability of home service robots,and has gradually become a key research content in the research direction of home service robots.This thesis takes the service robot independently developed by the research group as the research object,and uses the D-H parameter method to establish the positive kinematics model of the service robot's dual arms;On the basis of the positive kinematics model,the Monte Carlo algorithm is used to solve the cooperation space of the dual-arm system of the service robot;The inverse kinematics problem of the two arms of the service robot is analyzed using the Pieper algebraic solution method that satisfies the three-axis intersection.The inverse kinematics solving process program is written in MATLAB and the workspace of the two-arm cooperation is visualized and simulated.In the operation mode setting of dual-arm cooperation,two operation modes of asynchronous motion and synchronous motion are defined.In asynchronous motion mode,no constraints are imposed between the two arms,and master-slave planning is adopted.In the synchronous motion work mode,after adding constraint conditions between the two arms according to the position and posture information of the object,the motion trajectory of the main arm is prioritized.According to the trajectory and constraint conditions of the end effector of the main arm,the slave arm makes motion planning,so that the end effector of the two arms keeps relatively stationary.When dealing with the anti-collision problem in the process of arm motion planning,the detection method of AABB bounding box is adopted to simplify the arm model.Aiming at the shortcomings of redundant path sampling points,unsmooth path planning and poor security of the path planned by the traditional RRT-Connect algorithm in arm path planning,An improved algorithm is proposed,in which redundant nodes are deleted to simplify the path,parabolic transition is used to smooth the path in the path transition area,obstacles are surrounded by the ball model and the safe distance is calculated.The traditional RRT-Connect algorithm and the improved RRT-Connect algorithm are compared and simulated in MATLAB.The results show that the improved algorithm plans a simpler,smoother and safer path for the home service robot arm.In order to verify the feasibility of this algorithm,the URDF model of robot is visualized in ROS platform,and the simulation test environment of service robot dual-arm system is built.According to the two dual-arm cooperation modes,two different planning schemes are configured.Two simulation working environments with different working modes are built,and the motion planning and execution of the two operating tasks are completed.This thesis studies the coordinated operation of home service robot and proposes two cooperative strategies for service robot performing different tasks.This thesis focuses on trajectory planning during arm movement,and improves the traditional RRT-Connect algorithm for the working characteristics of the service robot.By adding the obstacle surrounding model,we define the safe distance,secure the home service robot,and then adopt the linear function smoothing path with a parabolic transformation.The proposed method is universal and extensive,which is not only applicable for the robot planning and design in the simulation environment,but also can have positive academic significance and application value in the extension of the planning and control of practical robots.