An Optimization on 2D-SLAM Map Construction Algorithm Based on LiDAR

When a mobile robot moves in an unknown environment, the emergence of Simultaneous Localization and Mapping (SLAM) technology becomes crucial for accurately perceiving its surroundings and determining its position in the environment. SLAM technology successfully addresses the issues of low localization accuracy and inadequate real-time performance of traditional mobile robots. In this paper, the Robot Operating System (ROS) robot system is used as a research platform for the 2D laser SLAM problem based on the scan matching method. The study investigates the following aspects: enhancing the scan matching process of laser SLAM through the utilization of the Levenberg–Marquardt (LM) method; improving the optimization map by exploring the traditional Hector-SLAM algorithm and 2D-SDF-SLAM algorithm, and employing the Weighted Signed Distance Function (WSDF) map for map enhancement and optimization; proposing a method for enhanced relocation using the Cartographer algorithm; establishing the experimental environment and conducting experiments utilizing the ROS robot system. Comparing and analyzing the improved SLAM method with the traditional SLAM method, the experiment proves that the improved SLAM method outperforms in terms of localization and mapping accuracy. The research in this paper offers a robust solution to the challenge of localizing and mapping mobile robots in unfamiliar environments, making a significant contribution to the advancement of intelligent mobile robot technology.