Research Of Slam And Navigation System Of Autonomous Rescue Robots Based On Ros

RoboCup Rescue is an international platform for researching and testing main technologies of robots which are used for rescue. In recent years, its development indicates that rescue robots will become smaller and more autonomous. Therefore, the research of autonomous robots with perception and intelligent decision-making ability is significant. Simultaneous Localization and Mapping (SLAM) and navigation technology is the key to autonomous robots.Based on the overall demand of automonous rescue robots, the design and implementation of SLAM and navigation system based on Robot Operating System (ROS) are presented in this thesis.First of all, the paper introduces the development of rescue robot and its related technologies. Then the RoboCup Rescue Robot Competition rules and the functional requirements of autonomous rescue robots are introduced. Finally the SLAM and navigation system is modularized.In SLAM modular, the robot pose is preliminarily corrected by scan matching algorithm based on grid map. Then accurate robot pose is acquired through Rao-Blackwellized particle filter. Finally we can build an incremental environment map based on accurate pose and sensor information.The navigation modular is designed for autonomous rescue robot to search disaster environment. Object points are generated based on the edge-detection algorithm. Then the robot obtains an optimal path from starting point to target point by global path planning algorithm, such as Dijkstra and A*. While the robot running along the path, the improved Dynamic Window Algorithm (DWA) is used to avoid dynamic obstacles.In this paper, we built the overall framework for SLAM and navigation system based on ROS and realized its corresponding functions. We used crawler robot to do experiments, deployed and tested the system in simulated disaster environment. Then we analysed the experimental results. The results show that the system is real-time and robust. It can finish the accurate positioning of robot, create precise and global consistent maps and complete explorating the environment effectively.