Study On Trajectory Planning And Its Control Methods Of Trajectory Following For Autonomous Trucks

Trucks have important practical value in the transportation of bulk cargo such as logistics and factories.However,the traditional trucks usually have fast driving speeds and long driving routes,which can cause the driver’s operational errors and greatly increases the probability of traffic accidents.Therefore,developing driverless truck technology is one of the most important ways to reduce the frequency of truck traffic accidents.Trucks usually have large turning radius,long body sizes and multiple axles,placing higher requirements on the optimal curvature of driving trajectory for the autonomous trucks.At the same time,the vehicle model is usually complicated in the tracking of the driving trajectory,which increases the difficulty of tracking accuracy.In view of this,the main research work of this paper is as follows:A local path planning algorithm for autonomous trucks(ego vehicle)in a structured road scenario has been proposed.In the scenario of free driving for the ego vehicle,the local path of the truck has been planned based on the natural cubic spline curve in this paper.Also,in order to make the generated trajectory as smooth as possible,the global optimization method is used,which is executed under the premise of giving a priority to the optimal curvature principle.In the scenario of actively avoiding obstacles,avoiding that the planned trajectory is unnatural and cannot meet the optimal curvature requirements of the driving trajectory,a “oval potential field” is introduced to construct a potential field area of the obstacle vehicles in front so that the scope of the potential field is adjusted,which is based on an improved artificial potential field method.Speed planning algorithms for the autonomous trucks in non-following and following cases have been presented respectively.In the free driving case,the sequential quadratic programming(SQP)method is used to plan speed for the ego truck.Meanwhile,a fourth-degree polynomial is introduced to describe the speed change of the truck during accelerating so as to ensure the continuity of the acceleration change of the ego truck.In the following case,the model predictive control(MPC)algorithm is used to plan the speed of the ego truck to maintain the distance between the ego truck and the following vehicle within an ideal range.The algorithms of speed and path tracking for the autonomous trucks have been proposed,and a co-simulation platform has been established to verify the effectiveness of these methods presented in this paper.First,a speed tracking controller based on the incremental PID control algorithm is designed to track the expected longitudinal speed of the truck in real time.Second,a path tracking controller based on six-degree-of-freedom vehicle model is also designed to improve its robustness and tracking accuracy.Finally,a co-simulation platform assembled by Pre Scan,Simulink and Truck Sim is used to verify the effectiveness of the speed and trajectory tracking control algorithms presented in this paper.