A robust planning and control system handling fuzziness

Abstract

A robust real-time planning and control system against the fuzziness, mainly caused by the behavior of the users, is proposed in this paper. The planner of the system constructs a tree-structured plan by a best-first algorithm giving priority to the reliable paths, and memorizes it. Based on the sensory information about the users, the controller executes the plan, selecting the most reliable and useful path of the tree. If no suitable path is available, the planner makes other plans within a limited amount of time. Therefore, the system realizes an interactive robot planning according to the user's reactions. To confirm the robustness of the planning system, we consider the simulation of an autonomous robot controlled by the planner, whose role is to assist a human worker. The robot transforms its plan according to the behavior of the worker.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>