Research On Path Following Control For Intelligent Vehicles

With the progress of society and the continuous development of the automobile industry,the huge car ownership has brought certain pressure on resources,transportation and the environment.In traditional driving behavior,the safety factor is the weakest link in the driver’s system.The emergence of smart cars has greatly alleviated the social problems caused by traditional cars.The research and development of smart cars is not only in line with the current trend of intelligentization of everything,but also an important guarantee for improving people’s wisdom and determining safe travel.Intelligent driving includes three aspects of technology,environment perception,decision-making planning and follow-up control.As part of the control layer,path tracking is a prerequisite for the realization of vehicle intelligence and practicality,and it is also the embodiment of the core value of intelligent driving technology.Therefore,it is of great significance to realize the path tracking control of fully autonomous unmanned vehicles under complex working conditions.This article mainly researches on the path following control of smart cars.First,a seven-degree-of-freedom dynamic model of the vehicle is established,and the Pacejka’89 magic formula is used to describe the tire model of the vehicle.Simulation analysis proves that the established model can meet the requirements of real-time and stability of intelligent vehicle path tracking control.A path following controller based on the optimal preview theory is proposed.The vehicle path tracking control problem is decoupled into the longitudinal speed following control problem and the lateral trajectory tracking control problem.The expected front wheel angle is converted into the optimal steering wheel angle input from the preview error model and the calculation of the heading angle deviation respectively,and the system’s lateral trajectory tracking controller is designed.Using a speed control system based on mode switching and a throttle/brake system to control acceleration input,the system’s longitudinal speed following controller is designed.In the preview tracking controller,PID feedback control based on preview error compensation is introduced,and a preview time adaptive module is added for the selection of preview distance.Simulation analysis verifies the tracking performance of the controller under three paths and different working conditions.The results show that it has a good tracking effect under high adhesion conditions,while the stability of the vehicle under low adhesion conditions is poor and the tracking accuracy is not high.A path following controller based on sliding mode variable structure control is proposed.Based on the bicycle model of the vehicle,a fuzzy-sliding mode vertical and horizontal coupling tracking controller is designed.The control input selects the function of the vehicle’s yaw rate,and the sliding mode surface is the vehicle’s yaw rate error.In order to weaken the chattering phenomenon of the sliding mode controller and ensure the control performance of the system at the same time,a fuzzy controller is designed to adjust the size of the sliding mode controller’s reaching law.Finally,the designed fuzzy-sliding mode path tracking control method is verified and analyzed under the co-simulation platform.The results show that the controller is suitable for vehicle driving at low and medium speeds and has high tracking accuracy.,The vehicle is prone to instability.A path tracking controller based on model prediction is proposed.The seven-degree-of-freedom dynamic model of the vehicle is selected as the model basis of the algorithm.In order to meet the requirements of real-time online optimization of the algorithm,a tracking control method based on linear time-varying model prediction is established,and an optimization objective function with the control increment as the state quantity is designed,and the objective function is transformed into a QP problem to ensure that the optimization objective is solved.At the same time,a soft constraint of slip angle is added to the design of the vertical and horizontal coupling controller to improve the vehicle’s path tracking performance.The designed model predictive controller was verified through simulation analysis,and the results showed that the model predictive controller with soft restraint of slip angle can adapt well to tracking driving under different vehicle speeds and different road attachment conditions.Finally,in the Carsim/Simulink co-simulation platform,the three path tracking control algorithms are simulated and compared.The results show that the preview controller is only suitable for the tracking of vehicles on high-adhesion roads,and the fuzzy-sliding mode controller is suitable for vehicles.Tracking driving at low and medium speeds has the highest control accuracy,and has strong adaptability to the extreme conditions of the road adhesion coefficient.The model predictive controller with soft constraints of the slip angle has unique advantages in the smart car’s response to the tracking problem of different vehicle speeds and different road attachment conditions.