| Road traffic accidents have produced severe effects on road safety performance, road capacity and efficiency. Moreover, frequently occurring traffic accidents have led to death and property damage of civilians. Vehicle collision is one of the main road traffic accident forms. The solution of how to effectively avoid vehicle collision accidents, and improve driving safety has been the common question to many scholars devoted to the field of Advanced Driving Assistant System(ADAS). In practical road traffic environment, vehicle collision accident involves multi factors, such as vehicle motion status, driver volition and behavior, road traffic environment(obstacles, road condition, etc.). Scholars have conducted in-depth research on vehicle intelligent collision avoidance technology. However, literature review in this paper indicates that many existed problems and challenges need to be solved to identify driving safety situation under complex road traffic environment, and make path planning strategy for vehicle collision avoidance under dynamic road traffic environment. The special issues include as following aspects: how to analyze the driving safety performance with comprehensively considering factors of driver volition and behavior, vehicle motion and dynamic road traffic environment; how to deal with the heterogeneity problem while fusing the data of “driver-vehicle-traffic”; how to build a situation assessment model to evaluate the driving safety under complex road traffic environment; and how to decide a safe collision-free path with considering the constraints, such as road traffic environment and vehicle dynamic characteristics.This study is supported by National Science Foundation of China “The identification method of vehicle collision threat based on driving intention and dynamic environment”, “Driver volition tracking and vehicle collision avoidance warning based on connected vehicle” and National High Technology Research and Development Program( "863" Program) of China “Key technology of connected vehicle-road driving safety system”. In order to resolve the problems existed in traditional vehicle collision avoidance system. In this paper, driving volition and dynamic road traffic environment are considered to study the methodology for identifying the vehicle conflict risk situation and making the collision avoidance path planning strategy under complex road traffic environment. In this paper, the main research work is as follows:1. The characteristics of driver behavior under typical driving volition(driving at constant speed, vehicle following, lane change) are analyzed. Then, the Linear Quadratic Regulator(LQR) method is applied to establish a driver behavior prediction model, which could quantitatively describe the driver's expected control manner under different driver volition pattern. Meanwhile, the proposed model can be used to simulate different driving habits by specify the model parameters.2. A novel vehicle motion prediction method is developed by considering both kinematical principle and real-time driver behavior. The improved vehicle motion model incorporates driver behavior into a constant acceleration(CA) model. Then, a quantitative approach is proposed to acquire the driver's expected control input based on a Linear Quadratic Regulator(LQR) optimal control method. In addition, Kalman Filter(KF) is applied to predict short-term vehicle motion, which is then used to analyze driving risks. Based on the trajectory prediction method, the driving volition and vehicle motion of multi vehicles within neighboring regions are considered to detect the conflict risk between vehicles, and a risk index is proposed to describe the conflict risk degree.3. A method is proposed for vehicle collision risk identification under the influence of the “driver-vehicle-traffic” multi-factors. The theory of situation awareness is applied to establish vehicle collision risk identification model with considering the fusion of related factors, such as driver volition, distance between vehicles, road condition, etc. Then, risk situation assessment algorithm is built based on theory of Variable Precision Rough Set. Similarity degrees between the current driving status and driving status in decision-making table are compared based on attribute weighted similarity, which could get the situation assessment results.4. An improved road potential field model is proposed for vehicle collision avoidance path planning based on the identification of driving safety. The improved model sufficiently considers the influence of driver volition and traffic environment on driving safety, and constraints of road boundary and vehicle dynamic characteristic to vehicle motion. Combined with the principle of grid algorithm, the traditional Artificial Potential Field model is improved to ensure that the vehicle motion along the planned collision avoidance path can effectively prevent collision accidents.Theory and methodology are studied to implement vehicle collision avoidance system under complex road traffic environment. On the one hand, the driving volition is considered within the identification of vehicle collision risk situation. By considering the driving volition, dynamic environment and vehicle movement state information to assess the current driving safety status, it can improve the identification accuracy of vehicle conflict situation risk. On the other hand, the vehicle collision avoidance path planning is studied by considering the road boundary constraints, factors such as vehicle motion state and driving volition are considered to propose an improved road potential field model, based on which, the vehicle collision avoidance path planning can more adapt to the complex road traffic environment. The proposed model and method can be used as theoretical basis to realize utility of vehicle intelligent collision avoidance under condition of cooperative vehicle-road system. Research results are of great significance to improve road traffic safety level, reduce the incidence of vehicle collision accident, and promote the development of intelligent transportation systems in China. |
PDF Full Text Request