Fuzzy-based Navigation and Control of a Non-Holonomic Mobile Robot

Abstract

In recent years, the use of non-analytical methods of computing such as fuzzy logic, evolutionary computation, and neural networks has demonstrated the utility and potential of these paradigms for intelligent control of mobile robot navigation. In this paper, a theoretical model of a fuzzy based controller for an autonomous mobile robot is developed. The paper begins with the mathematical model of the robot that involves the kinematic model. Then, the fuzzy logic controller is developed and discussed in detail. The proposed method is successfully tested in simulations, and it compares the effectiveness of three different set of membership of functions. It is shown that fuzzy logic controller with input membership of three provides better performance compared with five and seven membership functions.