Research On Key Techniques Of Autonomous Navigation Rescue Robot

Economic development of the.society consumes tremendous mineral resources,which come from the efforts of mineworkers.Mine disaster occurs around the world 'from time to time,this threatens the safety of mineworkers seriously.When cave mine disaster occurs,it is dangerous for the rescuers to survey the disaster field with little information,since other sub-disasters could happen or will hit the field again.Mine rescue robots can work effectively immediately when disasters happen for searching and rescuing people.A mine rescue robot can detect and find trapped people,and effectively rescue them.Conventional rescue robots utilize crawler.for moving,which rarely take into account of the road adaptability and intelligence when a robot is running on bumpy environments.Small suspension systems is proposed to:improve the stability of tracked robot to avoid big bumps.Meanwhile,the proposed rescue robot has two control schemes,autonomous navigation and manual teleoperation,which guarantee high efficiency in rescue tasks.This main work of this thesis is formatted as follows:..First,the survey,analysis and summary of current researches on rescue robots development around the world were carried out.Based on the analysis of advantages and disadvantages of the present rescue robots,this thesis proposes to design a rescue robot with higher road adaptability as well as autonomous navigation ability,which will be the main work of this study.Second,this thesis designed the hardware circuit of the robot control system,including carryon CPU selection,servo MCU selection,human detection sensor design,wireless transceiver module selection,wireless video module selection,and brushless motor drive design,toxic gas sensor detection circuit design,carryon battery system design,power circuit design,and remote controller system design.All the functional modules were tested independently,and the integrated system was tested eventually.ROS application framework method was employed here.Via this method,the SLAM for mobile robot can be developed much easier,and the program code can be designed individually,which improves efficiency.Meanwhile,navigation packages based on ROS can be easily transported to other mobile platforms,improving the sharing and reuse efficiency of the code.This thesis used VB to develop a user interface to display real-time environmental data within the disaster field,which provides first-hand information for the rescuers.Then,the robot movement modeling and error analysis were carried out.In this process,the PID control method was employed for robot motion control,which overcomes the "cannot go straightly problem" induced by the robot body transmission error and model error,and the corresponding experiment was conducted.In order to verify that the adaptability on roads with different conditions,robot motion simulation was conducted,including motion on the road with bosses,ramps as well as bumps.After simulations,the acceleration,velocity,and displacement curves in the vertical direction of the center of mass were shown and analyzed.Then,the two basic functions of SLAM,location and mapping,were programmed.The Carlo localization algorithm and Bayesian raster constructing algorithm were proposed,analyzed and programmed for SLAM through ROS.Experiments for verifying the robot localization and reliability were carried out via grid map building tests.The Dijkstra path planning algorithm was proposed for this rescue robot.The Dijkstra algorithm was analyzed and realized in ROS system.The trajectory planning tasks were conducted via the servo MCU Arduino,which receives the trajectory data from the host computer and drives the tracked robot autonomously.Finally,the robot hardware,software and mechanical structure were integrated,and the overall performance of the rescue robot was tested.Via the experiments,the following features of the designed robot are confirmed:maximum speed of 3M/s;it can easily climb 30°ramp;it is able to cross road with 300mm gully;also this robot can run on bumpy roads with gravel,or beaches smoothly with fast speed.Additionally,the robot is able to monitor and returns the video inside the mine cave in real time;it also can collect information about the concentration of toxic gases as well as temperature and humidity at the same time.In a dark environment,the rescue robot can light via a LED lamp and capture disaster field video.