Integrated Classical Planning and Motion Planning for Complex Robot Tasks

In order to execute complex tasks in complex environment successfully, robot need to make detailed planning for high-level task and low-level motion according to the requirements of task scenes. In this paper, a method based on the combination of classical planning and motion planning is discussed. In terms of high-level tasks, the robot decomposes a complex task into a collection of simple task, organizes simple task by logical reasoning and finally achieves the goal required to complete the task. In order to improve the adaptability of robot operation skills, the dynamic motion primitives were used to model the robot motion by imitating the interaction characteristics between human and objects. We set up a home service robot task scene to verify the feasibility of the proposed method.