A mobile robot relocation method based on dynamic region division

Abstract

The invention is applicable to the technical field of robot positioning control, and provides a mobile robot relocation method based on dynamic area division, including: loading a global map; determining a rough position of the mobile robot in the map, and determining the relocation based on the rough position area; generate random particles in the free area of ​​the relocation area, and limit the randomly generated random particles to the free area of ​​the relocation area; according to the position of the random particle in the global map, use the likelihood field to calculate the weight of each random particle; Screen out random particles with a large weight value, and add random particles at the same time; update the particle distribution, and update the maximum weight of the particle cluster. If the maximum weight of the particle is greater than the weight threshold, the average position of the particle cluster with the maximum weight is the best position. By dynamically selecting a relocation area with an appropriate size, it is ensured that the relocation area contains as much effective information as possible on the premise of the smallest area, and the relocation of the mobile robot is realized more quickly and efficiently.

Description

technical field [0001] The invention belongs to the technical field of robot positioning, and provides a mobile robot relocation method based on dynamic area division. Background technique [0002] With the development of science and technology, autonomous navigation of mobile robots is playing an increasingly important role in the fields of automated factories, intelligent warehousing and logistics, and household services. At present, the simultaneous localization and map construction (SLAM) technology of mobile robots has become a hot issue in robotics research, and is the basis and premise of autonomous navigation of mobile robots. There are methods based on kalman filter, method based on particle filter and method based on graph optimization to solve the SLAM problem. To realize the automatic navigation of the mobile robot, it is necessary to load the grid map generated by the mapping system, and realize the self-positioning and navigation of the robot on the map. Ther...

AI Technical Summary

Problems solved by technology

In the existing solutions to robot relocation, most of them realize robot positioning by pasting QR codes in the environment or installing auxiliary equipment such as UWB, which limits the scope of use of robots and increases costs; or uses particle-based The positioning method of filtering, that is, sampling the robot pose through the motion model to generate a large number of particles, updating the weight of the particles according to the observation results of the sensor and resampling, and continuously iterating to make the particles converge

In a large map, particle filter-based relocation converges slowly, and if the particle set does not cover the real pose well, the final particle will not converge to the correct pose

Images