Research On Vehicle Speed Adaptive Decision Control Considering Human-vehicle-road Coupling

Autonomous vehicles are an important part of intelligent transportation systems,which can effectively alleviate traffic congestion,reduce traffic accidents and reduce energy consumption.With the economic and social development of China,the demand for freight logistics is growing.In the commercial vehicle industry,especially freight vehicles,it will be the first area to achieve autonomous driving.As one of the most important systems of smart cars,longitudinal control is the ultimate function of automatic control of vehicle longitudinal motion,including speed decision and control execution.In this paper,a joint parameter estimation method based on the Kalman filterrecursive least square(KF-RLS)is established,which can accurately identify the key parameters of vehicles and roads.Meanwhile,a truck speed model considering driver’s behavior characteristics is established.An adaptive cruise control(ACC)system adapted to road geometry and target vehicle ahead is designed.The specific research contents and conclusions of this paper include the following aspects:Firstly,the joint parameter estimation method based on KF-RLS is designed.The designed estimator consists of two parts: a KF for filtering noisy states and an RLS for estimating parameters based on vehicle longitudinal dynamics model.The whole method is divided into three layers.The vehicle mass is estimated,and then the center of gravity(CG)position of the vehicle is obtained.On this basis,the road adhesion coefficient is identified.The simulation results show that the joint estimation algorithm has higher precision and anti-interference performance than single RLS.Secondly,the dynamic model of vehicle steering is analyzed,and the vehicle speed model is established from lateral and longitudinal safety.Moreover,the driver’s speed selection characteristics and the influence of the preceding vehicle on the speed of the vehicle are considered.By introducing the driver factor and the reaction headway,a speed decision model considering the coupling of human,vehicle and road is established.It provides a reference for speed decision-making of self-driving trucks.Finally,based on the previous research,an ACC system can adapt to road geometry and target vehicle ahead is designed.The whole system consists of two layers of controllers: the upper controller based on the model predictive control rolling optimization idea and the lower controller designed by the truck actuator.According to the different driving conditions,the control system is divided into three modes: free cruise mode,emergency braking mode and following mode,which can be automatically switched by the change of driving conditions.For each mode,the weight coefficients of the objective function and the corresponding constraints are readjusted to obtain the optimal control effect.Four typical working conditions under straight roads are used to verify the effectiveness of the multi-mode control system.Moreover,aiming at the accident-prone curve section of expressway,two groups of comprehensive driving conditions with different road adhesion coefficient are set up.The results show that the ACC system designed in this paper can make reasonable speed decision according to the road information and the target vehicle status,and can automatically switch different modes.It can take into account the multiple objectives of vehicle driving and achieve safe and comfortable driving,thus improving the performance of ACC system and expanding the scope of use.