Design And Implementation Of Autonomous Exploration System Of Rescue Robot

With the development of robot technology,the scope of research and application of mobile robots has begun to expand to all walks of life.In the scene of search and rescue after disasters,rescue robots have also begun to assist rescuers in environmental detection of disaster scenes,giving full play to their unique characteristics.Among them,autonomous exploration and mapping technology is one of the technical cores of rescue robots.Autonomous exploration and mapping refers to the rescue robot in the unknown post-disaster environment,without human manipulation,continuously generating exploration targets,moving to the target location,and during the movement,collecting environmental information through sensors,using simultaneous localization and mapping(SLAM)Algorithm to build a complete map model of the surrounding environment.At present,there have been rich research results in autonomous exploration of mapping technology,but there are still problems of reliability and practicability.Moreover,in the special application context of post-disaster rescue,robot autonomous exploration still lacks a mature solution.This subject is oriented to the background of post-disaster rescue,using crawler robots as the main robot to build a reliable rescue robot software and hardware system.The purpose is to design an autonomous exploration system suitable for search and rescue missions.The main research work of this paper is as follows:(1)Introduced the application of mobile robots in the field of disaster rescue,and analyzed the importance of autonomous exploration and mapping technology in rescue robots.Briefly summarize the research status and development direction of rescue robots,SLAM technology and autonomous exploration algorithms at indoor and abroad,and determine the specific goals of the subject research.(2)Analyze the movement model of the crawler rescue robot and determine the odometer model to calculate the robot's pose changes.The theoretical analysis of the SLAM algorithm in the autonomous exploration system,including Kalman filter algorithm,particle filter algorithm,Gmapping algorithm,and Hector SLAM algorithm,has laid a theoretical foundation for the selection of subsequent SLAM algorithm and the construction of rescue robot software system.(3)The hardware system of the rescue robot was designed and built,including the underlying driver panel,inertial measurement unit,lidar,adjustment gimbal,etc.Taking the ROS system as the robot software development framework,it is designed with the interaction layer,the decision-making calculation layer and the underlying driver layer,including the SLAM system,the autonomous exploration system,the underlying control algorithm,etc.,and introduces the main composition and advantages of the ROS system.A two-axis lidar adjustment gimbal based on an inertial measurement unit is proposed to ensure the ability of rescue robots to perform SLAM tasks in unstructured terrain.(4)Research on autonomous exploration methods of mobile robots,combined with existing autonomous exploration methods,propose a method for extracting exploration target points based on robot trajectory,and design exploration target point selection methods and path planning using gradient descent method.Comparative experiments on Gazebo physical simulation platform and traditional boundary-based exploration methods prove that the exploration algorithm proposed in this paper is more suitable for search and rescue missions.(5)Carry out real environment experiments on the autonomous exploration system of this paper,set up a flat ground and simulated post-disaster experimental site,and compare SLAM experiments on the lidar adjustment gimbal,which proves its effectiveness.Carrying out autonomous exploration experiments in two kinds of experimental sites proves that the autonomous exploration system built in this paper can complete and accurately map and explore the experimental sites without human intervention.