Comparison of Various Spacing Policies for Longitudinal Control of Automated Vehicles

Abstract

An automated vehicle controls its throttle and brake systems on the basis of the spacing policy, which is used to control the distance between vehicles. The spacing policy is required for maintaining a desirable state of vehicle movement in consideration of the movement of preceding vehicles. Because a spacing policy significantly influences the performance of automated vehicles, various spacing policies have been proposed. Previous studies have mostly focused on string stability, safety of the subject vehicle, and traffic capacity rather than on similarity to human driving behavior. However, developing automated vehicles with movements similar to vehicles driven by humans is important for automated vehicles that travel with manual vehicles driven by human drivers. In this study, the authors propose a risk-based spacing policy based on the microscopic analysis of the relationship between collision risk and driving behavior. The proposed spacing policy was simulated with real trajectory data and compared with the four existing spacing policies in vehicle operation characteristics, shock wave attenuation, and safety. To consider the effect of the market penetration rate of automated vehicles, the spacing policies were compared under the condition of an increased number of automated vehicles. The results show that the proposed spacing policy can significantly suppress shock waves. With the shock wave suppression, the newly proposed spacing policy can improve road safety, and its positive effect will grow as the market penetration rate of automated vehicles increases.