面向大范围场景的分布式激光惯性联合优化定位算法

A distributed LiDAR-inertial joint optimization location algorithm is proposed to address the problems faced by distributed platforms in large-scale complex environment location and mapping, such as dependence on satellite information and large positioning errors. Firstly, a no prior information location initialization method is proposed to address the issue of global coordinate unification across multiple platforms in satellite denial environments. Based on global features, a point cloud fusion algorithm is used to fuse local maps and achieve coordinate unification in dynamic environments. Secondly, a nonlinear optimization method based on cross platform global loop is proposed to address the real-time error accumulation problem in joint localization, achieving fusion localization of distributed sensor information. Finally, the effectiveness of the proposed method is verified through open-source datasets and vehicle-mounted experiments. The vehicle experiment in a large-scale campus scene of 3363 m shows that the proposed method can effectively improve the localization accuracy in large-scale and complex environments. Compared with the DiSCo-SLAM algorithm, the localization root mean square error and localization error standard deviation are reduced by 18.1% and 32.6%, respectively.