Safe Lane Change and Merging Gaps in Connected Environments ?

摘要

Humans sometimes choose less-than-safe lane change gaps, and this puts themselves and others at risk for brief periods of times. Autonomous vehicles supposed to operate in vehicle dense environments cannot replicate this risky behavior by design. On the contrary, they must guarantee collision-free operations under predefined worst-case scenarios. At the same time, overly conservative lane change gaps negatively impact traffic flow. In this paper, we adopt a worst-case scenario based on vehicles braking capabilities, and we use communications to decrease the safe vehicle following and lane changing gaps without compromising safety. The proposed safe gap is a linear function of the vehicle speed, which is an important feature in the design of stable and robust longitudinal vehicle following controllers. Simulations using realistic vehicle parameters are used to illustrate the approach.