Fuzzy control system for a mobile robot

摘要

Since the first fuzzy logic control system was proposed by Mamdani, many studies have been carried out on industrial process and real-time controls. The key problem for the application of fuzzy logic control is to find a suitable set of fuzzy control rules. Three common modes of deriving fuzzy control rules are often distinguished and mentioned: (1) expert experience and knowledge; (2) modeling operator control actions; and (3) modeling a process. In cases where an operator's skill is important, it is very useful to derive fuzzy control rules by modeling an operator's control actions. It is possible to model an operator's control behaviors in terms of fuzzy implications using the input-output data concerned with his/her control actions. The authors use the model obtained in this way as the basis for a fuzzy controller. The authors use a finite number of fuzzy or approximate control rules. To control a robot in a cluttered reactor environment, it is desirable to combine all the methods. In this paper, the authors describe a general algorithm for a mobile robot control system with fuzzy logic reasoning. They discuss the way that knowledge of fuzziness will be represented in this control system. They also describe amore » simulation program interface to the K2A Cybermation mobile robot to be used to demonstrate the control system.« less