Mobile robot positioning and mapping method

Abstract

The invention discloses a mobile robot positioning and mapping method, and relates to the technical field of robot positioning. The mobile robot positioning and mapping method comprises the steps of initializing the position of a mobile robot and a raster map of each resolution; estimating the possibility that the tail ends of laser beams in each raster map is occupied; iteratively solving the position change data, in the raster map with the lowest resolution, of the mobile robot from the last moment to the current moment; initializing a particle set; updating the position of the mobile robotin each particle; selecting a target particle and expanding the particle set; scanning the matching degrees between the particles in the particle set and the corresponding raster maps; calculating theweights of the particles in the expanded particle set; updating the corresponding maps according to the positions of the robot in the expanded particle set; selecting standby particles; and updatingthe raster map with the lowest resolution. The mobile robot positioning and mapping method is capable of solving the problem that the traditional particle filter-based SLAM systems are relatively large in positioning error and inaccurate in map due to wheeled speedometer errors.

Description

technical field [0001] The invention relates to the technical field of robot positioning, in particular to a method for positioning and mapping a mobile robot. Background technique [0002] Positioning and navigation is one of the prerequisites for mobile robots to realize intelligence. At present, many mobile robot simultaneous localization and mapping (SLAM) systems use grids to represent maps, and use particle filters to iteratively update robot position status and map status. [0003] The core idea of ​​particle filtering is to use more samples to represent all possible results of the state to be estimated, which is mainly divided into three steps: sampling, weighting, and resampling. The proposed distribution used for sampling mainly comes from the odometer estimation and lidar observation data of the mobile robot. Each sampling requires a lot of calculations to calibrate the odometer data. When the robot slips, collides, etc., this proposal It is difficult for the dis...

AI Technical Summary

Problems solved by technology

The proposed distribution used for sampling mainly comes from the odometer estimation and lidar observation data of the mobile robot. Each sampling requires a lot of calculations to calibrate the odometer data. When the robot slips, collides, etc., this proposal The distribution is difficult to accurately reflect the real position state of the robot, resulting in that the position represented by each particle in the particle set cannot match the map, and the system error is greatly increased

At the same time, the occurrence of resampling will also reduce the diversity of particles, making it difficult for the entire SLAM system to recover from wrong positioning information, resulting in tearing of the map. Operations such as path planning and obstacle avoidance based on this map will also be particularly difficult. difficulty

Images