Research On Autonomous Lane Changing Method And Trajectory Following Control Of Unmanned Vehicles

With the increasing demand for a good traffic environment and the increasingly close integration of high-tech and physical manufacturing,they has effectively promoted the development of intelligent vehicles.Highly unmanned driving can make people’s travel and lifestyle more intelligent,and effectively alleviate traffic congestion,promote driving safety,and therefore have good development and application prospects.Vehicle lane change is the most common driving behavior.Research data shows that among all the traffic accidents,most of them are related to the lane change operation.Therefore,this paper focuses on the decision-making mechanism,track re-planning and trajectory tracking control methods of unmanned vehicles,so that the independent lane change operation of unmanned vehicles is more safe,comfortable and efficient.The main research contents of this paper are as follows:(1)This paper first studies the decision-making mechanism of autonomous vehicles for independent lane change.By considering the driver’s personal driving habits,vehicle body structure and emergency braking distance influencing factors,an absolute safety minimum lane change quantitative model for unmanned vehicles under multi-discipline vehicle conditions is established.According to this model,the safe area and the non-safe area of the unmanned vehicle traveling are defined.The expected vehicle distance and speed dissatisfaction are used as indicators for the intention of the unmanned vehicle to change lanes,so as to make a specific decision on the lane change command of the unmanned vehicle.(2)The characteristics of the lane change trajectory are analyzed.The polynomial is used to fit the lane change trajectory,and the optimal lane change trajectory is selected by integrating the comfort of the lane change,lateral stability and road driving efficiency.Based on the dynamic environment,the search space is established,the obstacle vehicle trajectory is predicted,and the steady-state steering model is applied to generate the lane-changing feasible trajectory cluster,which is used as the NMPC prediction model,and the idea of rolling optimization is utilized.Combining the obstacle avoidance function cost function,the local expected trajectory lateral deviation cost function,the lane change stability cost function,and the related control quantity and control increment constraint,iteratively solves each trajectory planning window and selects the optimal trajectory point.(3)The geometric tracking method and model prediction method for trajectory tracking control are studied separately.Firstly,the principle of Pure Pursuit trajectory tracking control method is analyzed.Because the tracking accuracy is affected by factors such as vehicle speed and lateral deviation,Stanley control method is applied to compensate the tracking error of Pure Pursuit algorithm.Due to the acceleration of the lane changing conditions,this paper uses the incremental PID based on mode switching to control the longitudinal speed of the vehicle.In order to further improve the trajectory tracking effect and lateral stability of unmanned vehicles at high speed,design a trajectory tracking controller for the model predictive control algorithm,and develop an optimization objective function,and add dynamics and correlation the environmental constraints are iteratively solved.Considering the effects of vehicle model mismatch,time-varying characteristics,vehicle height nonlinearity and state deviation,this paper optimizes the model predictive control algorithm to adapt to the current system working environment.It can effectively avoid the system to generate iterative error affecting the traj ectory tracking effect.(4)The joint simulation platform was built in Matlab/Simulink and Car Sim to simulate the trajectory tracking control under double shift line conditions.The simulation results show that the improved Pure Pursuit algorithm has poor trajectory tracking ability under high-speed conditions.However,the MPC algorithm is more robust to speed changes.Finally,the designed decision mechanism,trajectory planning and adaptive MPC trajectory tracking controller are co-simulated.The performances of uniform lane change and accelerated lane change are verified respectively.The simulation results show that the overall algorithm can effectively realize all the functions of the multi-objective optimization set on the basis of ensuring the safety of the lane change of the unmanned vehicle.The autonomous lane changing and traj ectory following control method of the unmanned vehicle proposed in this paper has theoretical guidance and practical application value for the development of the driverless vehicle decision and control system.