Design And Implementation Of Path Planning And Control System For Driverless Low Speed Electric Vehicle

In recent years,driverless has become a hot topic in the society.Driverless plays an important role in transportation and logistics.Driverless path planning and control is the key to the breakthrough of driverless performance.Because most of the environment of driverless is at low speed,this thesis mainly studies the path planning and control method at low speed.The main contents of this thesis are global path planning,local path planning,horizontal and vertical control.For global planning,given the starting point and the ending point,an optimal path between the two points is found.The local path is to drive along the optimal path.Once the sensor senses the obstacle,the vehicle avoids the obstacle through the local path planning,so as to achieve the unmanned driving behavior between the two points.For the planned path,the control algorithm is used to control the planned path,so as to achieve a better control effect.The specific research contents are as follows:1)Aiming at the problem that the traditional A * algorithm can’t deal with Lane level information,a new A * algorithm based on Lane level high-precision map is proposed.Firstly,GPS is used to obtain the positioning information,and the geodetic coordinate is transformed into the plane rectangular coordinate system by UTM coordinate capture,and the high-precision map is generated according to the width of the lane.Then,taking the lane of high-precision map as the node and the line between lanes as the edge,the A * algorithm is used to obtain the optimal trajectory from the starting point to the ending point.2)In order to solve the influence of longitudinal motion on lateral motion at low speed,a lattice algorithm based on Frenet coordinate system is designed to decouple the whole path.Firstly,the global path planning trajectory is formed into a reference line,with the reference line as the center line,the path point is transformed into the point under the Frenet coordinate system,and the path is divided into the d-s coordinate system and the t-s coordinate system for solution;then the obstacle information,kinematic characteristics and dynamic characteristics are added to evaluate the trajectory;Finally,the lowest evaluated trajectory is synthesized,and the obtained trajectory is results of local path planning.3)In view of the problem of the accuracy of the longitudinal control and the lateral control,the longitudinal control adopts double PID to compensate the speed and acceleration,improves the acquisition of the calibration table,obtains the calibration table by using the current speed and control variables,and finally obtains the actual control quantity by fitting;the lateral control uses LQR algorithm to fine tune the control accuracy,and modifies the state matrix to obtain a good result Finally,the simulation of multi-level speed control shows that the control is feasible and the accuracy is high.