Dynamic Iterative Reconstruction Rescue Detection Intelligent Robot Based on ROS Platform

Yixiao Peng,Yuhang Li,Meng Na,Cao Feng
DOI: https://doi.org/10.1109/CAC59555.2023.10450203
2023-11-17
Abstract:At present, most rescue and detection robots are expensive, structurally complex, and have few functions. They only have modules such as tracking, obstacle avoidance, fire extinguishing, ultrasonic ranging, amphibious, and smoke detection. However, the use of these modules cannot perform deep detection in harsh environments and does not have predictive power. In order to make the intelligent robot carry out dynamic three-dimensional mapping of complex environment while rescuing, so as to better understand and analyze the environmental conditions and reduce the probability of dangerous events, a dynamic iterative reconstruction rescue detection intelligent robot based on ROS is designed. This design uses Nvidia's Jetson TX1 as the ROS master control, the ubuntu system as the carrier, and first uses the TOF ranging method and ROS of Lassen N10_ Cartographer carries out dynamic 2D reconstruction of the mine environment and scans the mine foundation, then uses Astra Pro to carry out dynamic iterative reconstruction of various environments with ORB-SLAM2 as the basic framework and scans the harsh environment obstacles to obtain the final environment, finally realizing the purpose of dynamic Iterative reconstruction of the environment by intelligent robots while rescuing.
Engineering,Environmental Science,Computer Science
What problem does this paper attempt to address?