Development And Implementation Of ROS-based Positioning And Navigation System For Self-balancing Fire Robot

With the development of robotics and fire protection industries,The requirements for related technologies of fire robots tend to be more intelligent.The positioning and navigation system of the fire robot reflects the environmental perception ability and independent thinking ability of the fire robot.The continuous development and improvement of the system is the basis of the intelligence of the fire robot and has important research significance.Based on the analysis and reference of the research status of fire robots and positioning and navigation technology at home and abroad,this thesis developed a self-balancing fire robots positioning and navigation system based on ROS system.The main work contents are as follows:1.According to the functional requirements and application occasions of the autonomous positioning and navigation system,a system development scheme with ROS system as the core that can communicate with the cloud server is designed,including the overall system design scheme and the selection of PC and sensor,etc;2.Study the principle of network communication,design ROS system and server network communication software,and customize communication protocol according to requirements.Study Linux system serial port programming principle,design ROS system and motion controller serial port communication software.The code was changed into the software form of ROS node,and the ROS Package was added to facilitate unified project management and data communication;3.In order to avoid EKF-SLAM's requirement on environmental noise,the FastSLAM algorithm based on RBPF was selected to decompose the localization problem and mapping problem in SLAM,and the optimal state estimation was carried out by particle filter and extended kalman filter respectively.A filtering method for covariance cross fusion is proposed.Finally,the FastSLAM algorithm is combined with the robot platform to complete the map construction function;4.Study and analyze the commonly used path planning algorithm,compare the advantages and disadvantages of different algorithms,choose A* algorithm for globalpath planning,Dijkstra algorithm for local path planning.The navigation function is realized on the robot platform with the network communication software,at the same time,for the dynamic balance motion performance of the self-balancing structure robot,the parameters of the path planner are debugged to achieve the best performance of the algorithm for the dynamic balance of the robot.Finally,the content of the full text research is summarized,the advantages and disadvantages of the system are analyzed,and several possible future research directions for the positioning and navigation system of fire robots are proposed.