Research On Transverse And Longitudinal Control Method Of Multi-Mode ACC System For Passenger Vehicle Based On MPC

With the continuous development of vehicle intelligence,adaptive cruise control(ACC),as one of the key technologies of advanced driving assistance system(ADAS),is gradually favored by people.At present,ACC system can only control the change of vehicle longitudinal speed.When the current vehicle speed is far lower than the expected speed and the surrounding environment meets the lane change,the vehicle cannot change lanes,which reduces the application scope of ACC system and passenger satisfaction.Therefore,it is of great significance to combine ACC longitudinal control and lane change control.Firstly,the layered architecture is determined to design the ACC system,and the ACC system is divided into constant speed,following,early warning obstacle avoidance,emergency obstacle avoidance and lane change modes according to the requirements of different working conditions.This paper points out the shortcomings of using distance scheme for mode switching,analyzes the factors affecting longitudinal mode switching and autonomous lane changing,and determines to use fuzzy logic theory.According to the influencing factors of mode switching,the longitudinal mode switching model and lane changing decision-making model are established respectively.Finally,considering that the fixed speed mode and follow mode take up a long time and have a high switching frequency,different switching transition schemes are designed by analyzing the vehicle performance requirements during the switching of the two modes.Secondly,according to the driver’s two operation modes of throttle and brake during vehicle driving,the inverse dynamic models that can obtain the braking pressure and throttle opening through the information of acceleration and vehicle speed are established respectively.Considering that the driver operates the accelerator and brake independently in the actual driving vehicle,when the engine is at idle speed and the vehicle runs at different speeds,the impact on the longitudinal speed of the vehicle is different,the switching strategy of accelerator and brake is formulated.On this basis,an incremental PID controller is added to adjust the output value of acceleration to ensure the driving stability of the vehicle.Then the control algorithm under each mode is established.For the following mode,the model predictive control algorithm is used to design.Considering the influence of the deviation between the MPC predicted value and the actual state value on the following accuracy,the prediction error compensation is introduced into the prediction model of model predictive control to improve the robustness of vehicle following mode.Because the fixed coefficient can not meet the performance requirements of vehicles in different following scenes,the weight coefficient is adjusted by fuzzy logic theory,so as to improve the following stability in following mode.For lane changing mode,MPC algorithm is used to design the trajectory planning layer and trajectory control layer.In the trajectory planning layer,in order to reduce the memory occupation of trajectory discrete points and improve the speed of tracking control layer to obtain reference points,the discrete points of planned trajectory are polynomial fitted and polynomial parameters are output.In the track tracking layer,the objective function is established according to the deviation between the predicted state quantity and the reference quantity and the control increment.The relaxation factor is added to the constraint to broaden the constraint boundary,ensure that the vehicle has a solution under complex conditions,and prevent the infinite extension of the upper and lower bounds of the constraint.The quadratic term of the relaxation factor is added to the objective function,so that the vehicle can deal with more complex scenes.Finally,the designed control strategy is verified by Car Sim and Simulink software.Five working conditions are selected for simulation analysis,and analyzed from the following,comfort,fuel economy and safety.By comparing the simulation results of improved MPC,unimproved MPC,PID and LQR control algorithms,the superiority of the control algorithm is verified according to the results.After the HIL scene is built,several typical working conditions are selected for test and verification.According to the experimental results,the control strategy designed in this paper has good effects in comfort,safety and follow-up.