Simultaneous Localization and Mapping Based on (μ+1)-Evolution Strategy for Mobile Robots

摘要

Simultaneous Localization and Mapping (SLAM) is one of the most important capabilities for autonomous mobile robots, and many researches have been proposed demonstrating the effective SLAM methods. However, these SLAM methods sometimes require assumptions such as the sensor model, which is difficult to implement and use the SLAM methods. In our previous work, a SLAM method based on Evolution Strategy (ES) was proposed and the on-line SLAM in indoor environments was realized. However, the definition of the map building method was not clear. Therefore, we propose a SLAM method based on a simple map building and search method. In this paper, we explain our autonomous mobile robot system and propose our SLAM method based on (μ+1)-ES. The experimental results show the effectiveness of the proposed method.