The Fuzzy Navigator with A Local Minima Solver for Real-time Self-reaction of A Mobile Robot in Flexible Manufacturing Cell/System (FMC/FMS)

XW Ma,YF Ma,HG Cai,O Khatib

DOI: https://doi.org/10.1109/icsmc.2005.1571756

2005-01-01

Abstract:This paper presents a novel two-layer fuzzy-navigator that incorporates sensing, control and planning to guide real-time self-reaction of a mobile robot in FMC/FMS. It can identify and solve the local minima problem during the robot's movement, which mimic the way that a human might understand his trapped state by recollecting some of similar experiences he had experienced before. The first-layer-controller does the regular navigation, while the remembrance is provided by the second-layer local- minima-solver, which can also analyze the information of sensor readings, impart an understanding of the robot's local environment and correlate the same with human heuristic experiences of a similar environment. The robot's local environment is identified in terms of fuzzy-rule, sensor- information and landmark-weight-vector. When the robot reaches a dangerous landmark-weigh-value during movement it understands its entanglement in a loop and takes suitable actions to pull the robot out of its trap. The simulation results prove the validity of the proposed method.