Affiliations: Chair of Computer Science Foundations, University of
Information Technology and Management, Rzeszow, Poland. E-mail:
[email protected] | Department of Electrical and Computer Engineering,
University of Manitoba, Canada. E-mail: [email protected] | Institute of Mathematics, Rzeszow University, Poland.
E-mail: [email protected]
Abstract: The Khepera robot belongs to the family of miniature mobile robots
of the K-Team firm. It is used in a number of places for scientific and
educational purposes. Considering its advantages (such as small size, precision
of movement, ease of control), it is applied to testing different approaches in
the domain of artificial intelligence. This paper describes the methodology of a control system design for
the Khepera robot based on a rough set approach. The proposed approach entails
a study of robot behaviour insofar as its movements are influenced by
measurements fromits sensors and the choice of actions that make it possible
for the robot to achieve its system goals. The constructed controller concerns
the realization of some tasks such as avoiding the obstacles, reaching a
target, following an obstacle, finding the way out of a labyrinth. The proposed
controller has been tested on both a robot simulator and on a real robot. Our
experimental results show that the proposed rough set methodology can be
applied to the design of a controller for the Khepera robot.
Keywords: Artificial intelligence, rough sets, fuzzy systems, machine learning, Khepera robot, control design, expert system
