Mobile robot navigation based on multisensory fusion

Abstract

Multisensory fusion is being increasing viewed as an important activity in the filed of mobile robot navigation and obstacle avoidance. The fusion of data from a variety of sensors makes the mobile robot more easily survival in a hostile environment. It takes advantage of the redundancy and reciprocity of multisensory data and increases the precision and reliability of inference and judgment for the mobile robot. This paper presents a method which employs fuzzy logic and neural networks to fuse data from several kinds of sensors. As a result, more exact navigation and quick obstacle avoidance can be achieved. © Springer-Verlag Berlin Heidelberg 2005.