A modeling methodology for car-following behaviors of automated vehicles: Trade-off between stability and mobility

Empirical studies have indicated that automated vehicle (AV) automakers tend to prioritize mobility over stability in designing car-following (CF) models, which may raise safety concerns. A likely explanation for this issue is that hardware-induced response delays challenge the ability of the CF models, as designed by automakers, to maintain an equilibrium between stability and mobility. To address these concerns, this study proposes a modeling methodology for the CF model in AVs aimed at achieving a trade-off between stability and mobility. This methodology seeks to identify the optimal parameters that enhance mobility under stability constraints. First, the linear CF model is calibrated using data from 20 commercial AVs produced by multiple automakers, and the unique response delay values of the linear CF model for each AV are identified. Next, the parameter regions ensuring stability are derived theoretically based on the calibrated response delays for each AV. An optimal mobility objective function is constructed to minimize time headway and reaction time, with the boundaries of the stable parameter regions serving as constraints. It allows the selection of CF parameters that maximize mobility while remaining within the stable regions. This proposed modeling method is applied to all AVs, and the optimal parameters are tested in simulations. Simulation results demonstrate that the proposed optimal model effectively dampens oscillations, reduces safety risks, and maintains shorter spacing, thus achieving an ideal trade-off between stability and mobility for AVs.