Base Pose Selection Of Home Service Robot Using Inverse Reachability Map And Operational Area Analysis

With the development of artificial intelligence technology,service robots are increasingly appearing in human life.In the home environment,service robots need to solve basic household tasks such as object classification and floor cleaning.The home environment has the characteristics of complex scenes and changeable objects.These challenges have added more difficulty and put forward more requirements for the programming of service robots.In the grasping process,the position of the robot to be reached mainly depends on the prior knowledge of the operator,and there is no standard for the best grasping position of the robot base.This thesis proposes a method based on the combination of the real-time operational area and inverse reachability map to complete the robot base pose selection in the grasping task.It provides a quantitative description for the selection of the robot base pose in the grasp task.The main contributions are summarized as follows:1.Use the inverse reachability map to select the base pose of the robot in the grasping task.This thesis conducts an in-depth study on the grasping performance of the Fetch robot by calculating the reachability index at different positions in the manipulator workspace and generating the reachability map of the robot.According to the reachability map,the inverse reachability map of the robot is obtained,which can be used to select the basic posture of the robot under the grasping task.Experiments prove that the grasping success rate under the robot base pose generated by the inverse reachability map has been greatly improved.2.Propose the concept of real-time robot operational area in the home environment.This thesis proposes the concept of a real-time operational area of a robot.Considering that the obstacles in the home environment have the characteristics of real-time changes,by studying the area that the robot can reach under the current position and combining it with the potential base pose calculated by the inverse reachability map,the robot can choose a better base pose to grasp the object.3.Application and verification on different robot platforms.This thesis analyzes the grasp performance and mobile performance of the expanded Turtlebot platform.By coordinating the movement performance and grasping performance of the Turtlebot,the lack of workspace of the manipulator is compensated by the movement of the Turtlebot base.In addition,this thesis generates different inverse reachability maps for different types of objects and designs different manipulation strategies.In the multi-object pick-up-place task in the home environment,Turtlebot has achieved good results.This grasp strategy is based on the actual grasping task of the home service robot.By combining the robotic mobility and manipulability,the potential base pose generated by the inverse reachability map is combined with the real-time operational area to select the final base pose for the robot under the grasping task.The Fetch robot and Turtlebot robot experiments prove the universality and versatility of this grasp strategy.Through the analysis of the inverse reachability map and the operational area,the robot can generate the most suitable base pose for the grasping task in the complex home environment.