Semantic SLAM service robot navigation method in indoor dynamic environment

Abstract

The invention provides a semantic SLAM service robot navigation method in an indoor dynamic environment. The semantic SLAM service robot navigation method comprises the steps of (1) semantically segmenting information and extracting feature points; (2) checking the movement consistency, removing dynamic points, and estimating the pose of the camera by using stable static feature points; (3) constructing a semantic octree map; (4) initializing the pose of the service robot; and (5) through a DWA local path planning algorithm combined with a grey wolf optimization algorithm, solving the optimalspeed trajectory and determining the service robot driving path... The problems that the cognitive ability of an existing traditional SLAM scene is weak, the positioning accuracy of the service robotis not high and the traditional DWA local dynamic planning is easy to fall into a local optimal position are solved; the DWA local path planning method combined with the grey wolf optimization algorithm is provided, dynamic obstacle avoidance is achieved in combination with semantic information, the cognitive ability of the service robot to the environment during advancing is guaranteed, more accurate speed information can be provided, and meanwhile safety and reliability are improved.

Description

technical field [0001] The invention relates to a semantic SLAM service robot navigation method in an indoor dynamic environment, relates to the field of autonomous navigation of service robots, and in particular to the field of semantic SLAM service robot navigation methods. Background technique [0002] With the continuous improvement of living standards in contemporary society and the continuous improvement of people's demand for quality of life, various intelligent service robots continue to emerge, and better research on service robots has become increasingly important. The mobile robot SLAM method is the basis for intelligent service robots. The early traditional SLAM methods were mainly based on the assumption of a static environment. In recent years, the emergence of artificial intelligence methods, especially machine learning methods, has significantly improved the performance of mobile robots in dynamic environments, and greatly enhanced the ability to understand m...

AI Technical Summary

Problems solved by technology

[0004] The technical problem to be solved by the present invention is to provide a semantic SLAM service robot navigation method in an indoor dynamic environment, which solves the weak cognitive ability of the existing traditional SLAM scene, which leads to the low positioning accuracy of the service robot and the partial dynamics of the traditional DWA. The planning is easy to fall into the problem of local optimal position. The DWA local path planning method combined with the gray wolf optimization algorithm is proposed, which combines semantic information to realize dynamic obstacle avoidance, guarantees the cognitive ability of the service robot to the environment when it travels, and can provide more accurate speed Information, increase safety and reliability at the same time, realize the pose estimation and effective planning of the route of the service robot in the dynamic environment, and also meet the real-time obstacle avoidance requirements of the service robot

Images