Eco-Lanes: A flexible lane management scheme for enhancing traffic efficiency and promoting eco-friendly commuting

Abstract

Existing high-occupancy vehicle lanes and other managed lanes typically cater to a single vehicle type, leading to suboptimal utilization. A new lane management scheme, termed eco-lanes (ELs), is introduced in this study to provide a more flexible operational framework, enabling authorities to designate various vehicle access types across different road segments. We conceptualize ELs as instrumental in enhancing traffic flow efficiency and promoting the adoption of eco-friendly commuting practices. To assess the quantitative impacts of ELs on commuters’ mode and route choices on network operations, a combined mode split and traffic assignment (CMSTA) model is developed, where a cross-nested logit (CNL)-based multi-mode passenger assignment model addresses route overlapping, and a nested logit (NL)-based mode split model accounts for the hierarchical choices of vehicle type (manual driving or auto driving) and commuting modality (driving alone or shared mobility). Formulated as an equivalent variational inequality (VI) problem in path-flow space with asymmetric link impedance functions, its equivalence is established. A new iterative solution algorithm integrating route-swapping, mode-swapping, and weighted adaptive self-regulating average techniques is developed to efficiently tackle the CMSTA problem. By reformulating the CMSTA as an equivalent nonlinear programming (NLP) problem and taking the multimodal stochastic system optimum (MSSO) as the objective, we further propose a mixed-integer nonlinear program (MINLP) for optimal EL network design. Numerical examples demonstrate the effectiveness of the EL scheme in mitigating congestion, not only optimizing network flow but also promoting eco-friendly mode shifts.