Application of Fuzzy Neural Network in Parameter Optimization of Mobile Robot Path Planning Using Potential Field

Abstract

This paper discussed a new mobile robot path planning algorithm. It analyzed the traditional path planning algorithm that employs artificial potential field and fuzzy logic, and then develop a new method combing artificial potential field, neural network and fuzzy Logic, which realized real-time obstacle recognition and smooth motion in dynamic environment. This algorithm improves the traditional artificial potential field performance, while reducing its drawbacks' influence effectively. Experiment and simulation shows this method can make the robot avoid obstacles effectively and avoid being trapped in "dead area".