Development Of Mobile Robot Simulation Platform Based On ROS And Research On Pedestrian Tracking

According to the development requirements of simulation platform and interactive experimental projects in mobile robot experimental teaching,aiming at the problems of complex operation instructions,difficulty in understanding experimental principles and low experimental participation in experimental teaching,this article aims to improve the quality of experimental teaching,combined with the development trend of the robot simulation platform on the open source robot operating system ROS,builds a mobile robot simulation platform based on ROS,and designs pedestrian tracking experiments.The main research contents are as follows:(1)The overall framework design of simulation platform.According to the characteristics of experimental teaching objects without relevant operating experience and the investigation of the development method of robot simulation platform,the design requirements of mobile robot simulation platform for experimental teaching such as simple operation,standard software framework,open control interface,and data visualization are analyzed,and puts forward the overall framework of mobile robot simulation platform based on ROS.In order to simplify the structure of the simulation platform and improve the stability of use,the simulation platform is divided into communication control module,state monitoring module,visual simulation module and quick instruction module based on the idea of modularization.(2)Development of mobile robot simulation platform based on ROS.The application development framework QT is used to develop each functional module in ROS environment.The communication control module is developed to realize the remote communication and motion control between the simulation platform and the mobile robot;The state monitoring module is developed to realize the real-time monitoring of the speed,visual field image and task execution state of the mobile robot;The visual simulation module is developed to realize the motion simulation of the mobile robot.The sensor data and operation parameters are displayed graphically.The user can send commands to the mobile robot through the visual interface to control the movement of the mobile robot;The shortcut instruction module is developed to realize the quick call of ROS instruction.(3)Pedestrian detection and tracking based on lidar.According to the distribution characteristics of the lidar point cloud,the hierarchical clustering algorithm is used to realize the point cloud clustering processing.A pedestrian detection method based on the leg model is proposed,which realizes pedestrian detection by extracting the characteristics of the pedestrian’s leg.The Kalman filter is used to perform motion filtering on the tracking process of the mobile robot,which improves the stability of the tracking process and realizes the following of the pedestrian by the mobile robot.(4)Mobile robot experiment.The kinematics model of the turnlebot3 mobile robot is established.The navigation experiment and pedestrian tracking experiment of the turnlebot3 mobile robot are designed based on the simulation platform.The function of the simulation platform and the performance of pedestrian detection and tracking algorithm are verified.The experimental results show that the simulation platform can simplify the experimental operation.The experimental results show that the simulation platform can simplify the experimental operation,reduce the difficulty of understanding the experimental principle through the graphical display of data,and effectively assist the development of experimental content.The pedestrian tracking experiment can improve the students’ participation in the experimental process through the interaction between the mobile robot and students.The mobile robot simulation platform and pedestrian tracking experiment designed in this paper can be compatible with various types of mobile robots,which promotes the application of mobile robots in experimental teaching.