Study On Simultaneous Location And Mapping With Navigation Of Rescue Robot

In recent years,with the development of artificial intelligence technology,the environment of mobile robots has become more widespread.Among them,the more popular mobile robot's simultaneous positioning and construction(SLAM)technology can effectively improve the robot's ability to complete tasks autonomously.The research of mobile robots for disaster relief has evolved into a more intelligent one due to the progress of SLAM technology.This article designs a rescue robot that is used in a room after a fire.Its main purpose is to quickly create an environmental map,and at the same time feedback video,audio,temperature,and other information to the client to assist rescue workers in their work.Firstly,the research background and research significance of robot SLAM technology and the research status of rescue robots in China and abroad are described.Secondly,the theoretical problems in the use of rescue robots are studied and a rescue robot system model is established.Including laser range finder model,robot motion model and measurement model.Through the concept of the kinematic model and the observation model of the mobile robot,it provides the theoretical basis for the subsequent platform construction.Thirdly,the SLAM algorithm is studied.The positioning of the robot and the map construction adopt the gMapping algorithm based on Rao-Blackwelized particle filter and the scanning matching algorithm to improve the ICP-2D algorithm suitable for two-dimensional laser data.The environment is represented by a grid map.Then based on the ROS platform,a four-wheeled skid-steer drive rescue robot is designed and its URDF model is established in the Gazebo environment.Finally,the rescue robots were tested in an ordinary indoor environment and a mine environment.The gMapping and ICP-2D algorithms are used in the simulated Gazebo environment to compare the building effects under different vehicle speeds in an ordinary indoor environment.Experiments show that the robot uses the gMapping algorithm to build the maps more accurately when running at low speed,and can only use the ICP-2D algorithm to quickly build maps when driving at high speeds.In an ordinary indoor environment,experiments are performed on multi-robots to build maps simultaneously.Experiments show that the multirobot_map_merge package can be used to achieve multi-robot map fusion to improve overall construction efficiency.Experiments in the mine road environment have shown that RBPF has a very prominent effect on low-speed mapping in long straight roads,while the highspeed ICP-2D algorithm has high requirements for laser range finder and construction operations.After the overall comparison,the ICP-2D algorithm is more suitable for rapid robot building.After completing the construction,robot can return to the starting point through navigation,and can guide the rescue firemen to reach the trapped person to complete the rescue work.