Research On Path Planning Of Smart Car Based On Multi-objective And Dynamic Environment

The development of intelligent networked vehicle technology can solve the problems of traffic safety,road congestion,and energy consumption faced by the automobile society,and it is also the main body of building intelligent transportation.ICV refers to the installation of sensors,controllers,actuators and other equipment,integrating modern communication technology and network technology,with complex environment perception,intelligent decision-making,and automatic control functions,enabling information sharing and control coordination between vehicles and external nodes.The result is the next generation of vehicles that drive safely,efficiently and economically.The miniature smart vehicles in this to study are often called smart cars.This study takes indoor miniature smart vehicles as the research object to carry out path planning research.According to the establishment of the kinematics model and odometer model of the miniature intelligent vehicle,the posture state change of the miniature intelligent vehicle is derived,and the distance information of the obstacle scanned by the lidar to the world coordinate system is obtained through the coordinate transformation used by the miniature intelligent vehicle.In the part of synchronous positioning and map construction,the Gmapping algorithm based on filtering and the Cartographer algorithm based on graph optimization are analyzed,and the two SLAM algorithms are theoretically analyzed.After comparing the two mapping algorithms,the Cartographer algorithm is selected to build the map.The performance of the Cartographer algorithm is analyzed and compared for two different environmental conditions: the crowded indoor environment and the long straight corridor.Experiments show that the Cartographer algorithm can effectively build environmental maps.In the path planning part of the smart car,in the global path planning algorithm,the A* algorithm and the Dijkstra algorithm are theoretically analyzed and compared,and it is found through experiments that the search efficiency of the A* algorithm is faster than the Dijkstra algorithm.In local path planning,MATLAB simulation experiments are performed on the parameters of the evaluation function in DWA,and a set of relatively ideal weight parameters are obtained.Aiming at the difficulty of obstacle avoidance for small cars in small areas in the practical application of the algorithm,the Bug2 algorithm is proposed to improve the DWA algorithm.Finally,Autolabor Pro1 is used as the moving chassis of microfilm intelligent vehicle in this study.Sensors such as two-dimensional laser radar and inertial navigation are installed before and after the moving chassis to build an experimental platform for SLAM and multi-target path planning experimental demonstration.Experiments show that the path planning algorithm in this subject can be completed in a multi-objective and dynamic environment.