Safety-aware vehicle-following driving optimization of intelligent and connected vehicle at signalized road intersection

Abstract

The driving safety at signalized road intersections is a critical issue for intelligent and connected vehicles (ICVs). Vehicle-following driving scenarios are common conditions at road intersections, as the traffic congestion often occurs at these locations. This paper proposes a safety-aware vehicle-following driving optimization strategy (SAVFDOS) for ICVs at signalized road intersections. The framework of the SAVFDOS includes three layers, i.e., the situation assessment layer, the decision-making layer and the speed planning layer. The situation assessment layer evaluates the likelihood of the vehicle passing through the signalized road intersection and the safety of the vehicle-following driving. The decision-making layer determines the ICV's actions, such as acceleration, deceleration, cruising and stop. The speed planning layer outputs the planned speed based on the situation assessment layer and the decision-making layer. With the SAVFDOS, the rear-end collision and the dilemma zone (DZ) problem can be simultaneously avoided. The validations are conducted by comparing the proposed method with an advanced benchmarked method. Compared with the benchmarked method in the simulation scenarios, the average time proportion of the following vehicle in DZ by the proposed method can be decreased by 33%. In addition, the real-world validations demonstrate that the average time proportion of the following vehicle in DZ by the proposed method is lower than 25%. The validation results prove that the proposed method can eliminate the potential risks of ICVs in vehicle-following driving scene at signalized road intersection.