A vehicle in a platoon sometimes faces the risk of causing a rear-end collision when it is following a vehicle in front. In the stability theory of the well-known GM car-following models, it has been said that the fluctuations of vehicle speeds and headways will increasingly propagate to the rear vehicle in a platoon if the platoon is asymptotically unstable. However, almost no research has been done to validate this phenomenon with the real car-following platoon data. Therefore, we used the car-following platoon data including ten vehicle trajectories to evaluate asymptotic stability in the platoon. We first tried to evaluate the asymptotic stability in vehicle safety because we assumed that the vehicle in the rear position isexposed to riskier conditions than the vehicle positioned in front due tothe shorter headways created by increased fluctuations of speeds and headways in a platoon if the asymptotic stability is unstable. We defined three indicators for safety: potential danger time, impact speed, and expected impact speed. Our outcomes in these safety indicators did not show the car-following platoon was in asymptotically unstable conditions. Therefore, we defined two supplemental indicators to closely observe asymptotic stability: maximum speed amplitude and maximum spacing amplitude. With these stability indicators, we were able to explain that the car-following platoon was under asymptotically unstable conditions. As a result, a hidden aspect of the relationship between asymptotic stability and vehicle safetywas discovered in the real car-following platoon.
Abstract