Robot positioning method

Abstract

The invention discloses a robot positioning method. A to-be-positioned robot is accessed to a wireless local area network on the basis of WiFi. The robot positioning method comprises steps that estimated present location positions of the robot are determined based on WiFi fingerprint positioning; a map of the estimated positions is called, the constraint estimated positions of the map are marked, and the precise position of the robot is acquired by employing a particle filtering algorithm. The robot positioning method is advantaged in that positioning precision is relatively high, and the required iteration time is relatively short.

Description

technical field [0001] The invention relates to a robot positioning method. Background technique [0002] With the development of science and technology and the improvement of people's living standards, for example, service robots have gradually entered people's lives, providing various humanized services for humans. Compared with other types of robots, the working environment of service robots is more complex, so autonomous and intelligent positioning in complex environments is a prerequisite for service robots to provide high-quality services to humans. [0003] At present, the positioning problems of service robots are mainly divided into two types: position tracking and global positioning. Among them, global positioning refers to the determination of the pose of the robot through sensor perception information without any prior knowledge of the initial pose. Global positioning is mainly used to solve problems such as robot position initialization, robot kidnapping, and ...

AI Technical Summary

Problems solved by technology

[0008] Due to the high complexity and dynamics of home and office environments, the global particle filter positioning algorithm has the following problems in such environments: 1) Due to the large map and the large number of particles required, the calculation time of global particle scattering is relatively long ;2) When the sampled particles are close to the real position of the robot, the success rate of the particle filter positioning algorithm can be guaranteed, but when the map is too large and the number of particles is limited, the sampled particles cannot cover the real position of the robot, resulting in a decrease in positioning accuracy; 3) In an environment with strong structural similarity, the success rate of global positioning is not high

[0009] In response to this problem, the literature "Global Positioning of Mobile Robots Based on Monocular Vision" (Robot, 2007, 29(2): 140-144, 178. Li Maohai, Hong Bingrong, Luo Ronghua, etc.) uses monocular vision to create an environmental map and pass Feature matching realizes global positioning; however, image processing is more complicated, and the monocular vision system increases the cost and calculation burden of the robot

The literature "Research Progress in Localization and Map Creation of Mobile Robots Based on Particle Filters" (Robotics, 2007, 29(3): 281-289, 297. Yu Hongshan, Wang Yaonan) introduces some methods of robot map creation and localization using particle filters , but in a complex environment due to the similarity of the scene structure, a single particle filter localization is likely to cause localization failure without any prior knowledge of the robot position

Images