This paper focuses on localization and mapping issues for autonomous mobile robots equipped with low-cost 2D lidar in complex environments. Most existing solutions commonly parameterize the robot pose on SE(3) when the robot moves on the rough ground and uses the scan data that may be insufficient or sparse to build the map. In this paper, we first developed the Gaussian Process (GP) to address insufficient scan data for low-precision 2D lidar by enriching the lidar measurements at interest or specific bearing regions. Meanwhile, A new method, based on the graph optimization framework, to solve the non-SE(2) perturbations is proposed, namely SE2-3D constraint, which directly parameterizes the robot pose as SE(2) without ignoring the non-SE(2) perturbations by associating the extended SE(2) pose with map point via lidar measurements. The experimental results indicate that the raw lidar data processed by our method can generate higher quality maps than the original data under the same working conditions. The simulation results verify that the proposed method has higher performance in terms of accuracy than traditional methods. This paper provides a meaningful solution for the broad application of ground mobile robots equipped with low-cost 2D lidar.

本文重点研究了在复杂环境中配备低成本 2D 激光雷达的自主移动机器人的定位和映射问题。当机器人在粗糙的地面上移动时，大多数现有解决方案通常在 SE(3) 上参数化机器人位姿，并使用可能不足或稀疏的扫描数据来构建地图。在本文中，我们首先开发了高斯过程 (GP)，通过丰富感兴趣或特定轴承区域的激光雷达测量来解决低精度二维激光雷达扫描数据不足的问题。同时，提出了一种基于图优化框架解决非SE(2)扰动的新方法，即SE2-3D约束，它直接将机器人位姿参数化为SE(2)而不忽略非SE(2) 2) 通过激光雷达测量将扩展的 SE(2) 姿态与地图点相关联来进行扰动。实验结果表明，在相同的工作条件下，我们的方法处理的原始激光雷达数据可以生成比原始数据更高质量的地图。仿真结果验证了所提出的方法在精度方面比传统方法具有更高的性能。本文为配备低成本二维激光雷达的地面移动机器人的广泛应用提供了有意义的解决方案。