Experimenting fuzzy control strategies for mobile robots on a rapid prototyping platform

This paper describes a rapid prototyping platform developed to address the complex design of real-time decision and control strategies for commercially-available mobile robots. The platform consists in a microprocessor board fully controllable in Matlab/Simulink object-oriented programming environment, directly connected through a serial link and a specific communication protocol to the mobile robot. The architecture also encompasses a vision sensor that can be used as a further source of feedback for real-time control, allowing the users to develop, test and compare directly on the hardware navigation strategies of considerable complexity. To show the practical advantages of the platform, this paper illustrates the design of an effective behaviour-based fuzzy control strategy guaranteeing simultaneously obstacle-avoidance and targetreaching performance in an unknown environment