Research On Scene Perception And Path Planning Algorithm Of Mobile Robot

Scene perception and path planning are indispensable components in mobile robot technology.The role of scene perception is to detect obstacles in the environment,while the role of path planning is to plan a collision free and optimal path from the current location to the target location under known environmental conditions.Traditional environmental perception and path planning methods cannot meet the current performance requirements of robots in society.This article makes some improvements and has certain potential application value in indoor service robots.The main research content includes the following aspects:1.A indoor environment perception method based on the fusion of single line Li DAR and depth camera is proposed to address the issues of susceptibility to interference and incomplete information collection when using a single sensor to perceive the external environment in current indoor scenes.Firstly,convert the infrared data collected by the depth camera into point cloud data,then perform voxel filtering on the point cloud to reduce the number of point clouds,and use a plane segmentation method to remove the ground plane point cloud.Using deep learning networks to recognize color images captured by depth cameras,classify obstacles,and finally fuse the information detected by depth cameras and single line lidars using an extended Kalman algorithm.2.A navigation obstacle avoidance method that combines the improved A *algorithm and the improved dynamic window algorithm is proposed to address the problems of multiple inflection points,long time consumption,and low security and flexibility of the dynamic window algorithm.Construct a two-layer cost map with different granularity values,and use the A * algorithm to plan global path key points in the two-layer map.Expanding the safety and comfort zones of different sizes around different categories of obstacles in the cost map,and expanding the evaluation function of the dynamic window method,the robot has different avoidance distances for different obstacles,and adjusts its movement speed based on its distance from the obstacles.Finally,the global key inflection point is used as a temporary target point for local planning to achieve the fusion of the two algorithms.3.Finally,simulation comparative experiments were conducted in the Matlab environment,and a mobile robot platform was built based on the ROS operating system to conduct obstacle detection and path planning experiments in indoor environments.The experimental results show that compared to traditional methods,the improved environmental perception method in this paper is more accurate and comprehensive in detecting obstacles;The improved fusion path planning algorithm not only reduces the number of global path folds,but also improves the flexibility,security and mobility efficiency in the local dynamic programming process.