Algorithm Design of Environment Map Construction for Inspection Robots Based on Data Fusion

Inspection robots are a hot research topic in the field of mobile robots, and the construction of environmental maps is a prerequisite for robot inspection. However, the existing LIDAR-based map building algorithms can cause inaccurate measurements due to distortion and dithering during operation. In this paper, a loosely coupled LiDAR-IMU fusion scheme is designed at the front which can processing the pre-integration information of the IMU, then the distortion of motion in the laser point cloud will be remove, thus, the alignment accuracy of two adjacent point clouds is improved. The back-end uses a tightly coupled fusion approach to obtain the positional information based on error-state Kalman filtering by jointly optimizing the residual terms of the IMU with the residual terms of the laser point cloud matched with the local map. The results are tested on the open-source dataset KITTI, and the environmental maps with better performance are obtained.