Research on mobile robot indoor positioning mapping based on front-end and back-end optimization

When the traditional RBPF-SLAM algorithm is applied to indoor positioning and mapping of mobile robots, it is prone to inaccurate positioning and mapping due to the long computation time of the front-end matching algorithm and the presence of particle missing in the back-end optimization algorithm. We popose to optimize the matching efficiency of lidar using the PL-ICP algorithm in the front-end of the SLAM system. In the backend of the SLAM system, the RBPF-SLAM algorithm is optimized using an improved beetle swarm algorithm. The experimental results show that the PL-ICP algorithm can significantly reduce computation time in front-end matching. In terms of backend optimization, the improved RBPF-SLAM algorithm can significantly reduce the number of particles and resampling, effectively improving the accuracy of indoor positioning and mapping for mobile robots.