Self-Organizing Neural Networks for Simultaneous Localization and Mapping of Indoor Mobile Robots

Abstract

Since the complexity and variety of indoor environments, the ability of simultaneous localization and mapping for autonomous mobile robots restricted their applications. A novel approach, which is based on clustering algorithm, fuzzy logic and neural networks, is proposed, and solve simultaneous mapping and localization problem. It adopts self-organizing fuzzy neural networks to model the environment and to implement localization. The neural map can be built on-line automatically. Meanwhile, the neural network exports robotpsilas pose information.