Research on Robust Localization Method for Coal Mine Robots Based on Static Map Construction and Multi-source Information Perception
Abstract：To address the issues of complex terrain, variable dynamic obstacles, and unstable GNSS signals in coal mine ground environments, this paper proposes a robust localization method based on multi-sensor fusion and hierarchical dynamic point cloud removal. A static high-precision map is constructed using multi-resolution depth image comparison and an improved ERASOR algorithm, effectively filtering out dynamic interference point clouds. On this basis, an environment-adaptive localization and re-localization framework is established by fusing adaptive Monte Carlo localization, IMU, and GNSS sensor information through an extended Kalman filter. The robot's abnormal motion detection and pose recovery are achieved through laser radar matching and GNSS status monitoring mechanisms. Experimental results show that the method achieves a dynamic point cloud removal rate of 78.1% in semindoor environments and has good re-localization response capability, making it suitable for applications such as coal mine inspection and transportation.
Machine-generated Keywords：
Publication Date：2025-11-30
Online Publishing Date：2025-12-10(First online date of this platform, not the publication date of the document)
Pages：10( 31-40 )