Fuzzy discrete event system based behavior modulation in mobile robotics

This paper presents a novel behavior modulation technique using fuzzy discrete event system (FDES) for mobile robots. The algorithm exploits the multi-valued feature of fuzzy logic (FL) and event-driven property of discrete event system (DES). Desirability (activity) of a behavior is expressed in terms of a fuzzy state vector. State-based prediction of the desirability is accomplished using fuzzy event matrices. A central arbiter employs priority based arbitration among the desirability state vectors and recommends new event matrices to modify the desirability states of the behaviors. The algorithm combines aspects of both command fusion and behavior arbitration. Furthermore, the proposed approach constitutes state-based observability and controllability to cope with the uncertainties of sensory data and dynamic environment. Simulation results of FDES-based mobile robot navigation are also presented to validate the performance of the proposed algorithm.