Research On Ackerman Car Trajectory Optimization Based On A-STAR Algorithm

With the rapid development of science and technology,the rapid rise of artificial intelligence,the popularization and application of big data,the intelligent research of mobile robots has set off a boom.At the same time,path planning is also the key to the autonomous movement of mobile robots in unknown environments.This paper focuses on the mobile robot based on Ackermann type steering gear,carries out the research on path planning,builds the Ackermann car platform,and designs the autonomous navigation system of the car.First of all,this paper introduces the overall structure of the experimental design of Ackermann car,the selection of the parts needed for the experiment and the reasonable layout of all parts of the circuit connection,and finally made into a real object.According to the structure of Ackermann trolley,the kinematics model and dynamics model are analyzed,and the relation expressions of wheel speed,Ackermann Angle and kinematics are obtained.Through ADAMS simulation and analysis of three kinds of motion states,the expression of the relationship between steering gear deflection Angle and front wheel rotation Angle is obtained.Secondly,the control system of Ackermann car is built.It mainly revolves around three contents: bottom control,top control,bottom and top coordination control.Among them,the underlying control is mainly PID speed regulation,and the use of APP for Bluetooth control;The top-level control is mainly the remote control of the PC to the IPC,and the IPC to control the radar under the ROS.The bottom level and the top level coordinated control respectively introduce the data format that STM32 and IPC receive and send each other,as well as Ackerman car's two ways of moving control under ROS.Finally,Robot?pose?ekf algorithm is used to integrate the IMU data with the odometer data,so that the Ackermann car can get a more accurate pose.Then,path planning of mobile robot is introduced in detail from global and local path planning respectively.The global path planning uses the A-star algorithm,which introduces the principle of the algorithm,and does the "trimming" processing to the algorithm.For A-star algorithm,the second derivative of minimum acceleration and the third B-spline curve are respectively used for trajectory optimization,and the former with better effect is selected as the trajectory optimization of A-star algorithm.The principle of Teb?local?planner is introduced,and static and dynamic obstacles are tested respectively in RVIZ simulation environment.By testing the local paths in these two states,the optimal local paths are obtained respectively,thus verifying the superiority of Teb?local?planner.Finally,according to ROS system,Gmapping algorithm and Move?base node are used to build the navigation framework.Through simulation and experimental verification: under this navigation framework,the Ackermann car designed in this paper can complete the autonomous navigation task from the starting point to the target point by combining with its own control and using the Teb?local?planner and the global path with the smooth and optimized trajectory of the A-star algorithm,and avoid objects in real time in the process of movement.