3D LiDAR SLAM for Roadheader in Coal Mines

The traditional UWB-IMU (ultra-wideband, inertial measurement unit) positioning method has been used in the positioning of roadheader, but its accuracy in the Z -axis direction and yaw angle is insufficient. In order to improve the localization accuracy, a 3D LiDAR SLAM (simultaneous localization and mapping) scheme is proposed in this paper. Through the registration of roadway point cloud, the pose of UWB-IMU is corrected to improve the accuracy. According to the work flow and complex environment of roadheader, a 3D scanning system with wide pitch angle is constructed and an improved cloth filter algorithm is proposed to segment the point cloud of the roadheader working area. In order to avoid few features affecting the matching results in the roadway environment, this paper adopts the way of the point cloud feature extraction based on curvature and uses the measurement results of UWB-IMU as the initial value for registration. The test results in the experimental roadway show that the positioning accuracy of the proposed method in Z direction and yaw angle is significantly improved compared with the traditional UWB-IMU.