Experimental Study on mapping and localization algorithm of intelligent wheelchair in spacious and dynamic environments

Abstract

As a tool to serve the elderly and disabled people, intelligent wheelchair may work in spacious and dynamic environments, such as parking lot. One difficulty of working in such scenarios is spaciousness and large scale which increases the difficulty of mapping. And the other is that there are various dynamic obstacles with different mobile frequency in the environment, which poses a new challenge to localization. In this paper, a multilayer matching based incremental mapping algorithm is designed to keep map accuracy and consistency in large scale and spacious environments and a localizability-based particle filter localization algorithm is utilized to maintain localization accuracy in dynamic environments. Experiments in two different parking lots verify the effectiveness of the mapping and localization algorithm.