Robot multi-sensor fusion positioning method and application thereof

Abstract

The invention relates to a robot multi-sensor fusion positioning method and application thereof in an intelligent cleaning robot, and the method comprises the following steps: installing a laser scanner and an IMU on a robot, installing encoders on two wheel shafts of the robot, and installing a GPS if the robot is outdoor and needs to use the GPS; if the UWB is required to be used indoors, installing the UWB, and installing the RFID and the camera; constructing an environment map of the place by the robot through the SLAM; during positioning, firstly, giving an initial pose of the robot, in the subsequent positioning process, using a graph optimization method to calculate the moving distance of the robot, wherein the moving distance is made to most accord with the moving distance, observed by all sensors, of the robot, and obtaining accurate positioning of the current robot in combination with the pose at the last moment; according to the invention, the current position can be accurately identified, the positioning accuracy is greatly improved, and the phenomenon that the robot loses positioning in an environment with poor map quality is avoided.

Description

[technical field] [0001] The present application relates to the field of intelligent robots, in particular to a multi-sensor fusion positioning method for robots. [Background technique] [0002] Positioning and navigation technology is one of the key technologies of robots, and positioning is the premise of navigation. Positioning means that the robot obtains its own position on the map by reading sensor information and matching it with the current map. In the SLAM problem, it is the most basic and important function of the robot to use the information perceived by the sensor to achieve reliable localization. [0003] However, robot application scenarios are complex, and sensors need to be changed accordingly for different scenarios, resulting in the uniqueness of positioning methods in each scenario, making the positioning methods less versatile. In addition, since most of the current positioning methods use the last positioning information to estimate the current positio...

AI Technical Summary

Problems solved by technology

[0003] However, robot application scenarios are complex, and sensors need to be changed accordingly in different scenarios, resulting in the uniqueness of the positioning method in each scenario, making the positioning method less versatile

In addition, since most of the current positioning methods use the last positioning information to estimate the current positioning information, and then correct it according to the current sensor observation data, so that the historical positioning information cannot be fully utilized, so it is easy to lose positioning.

Images