Multi-band model for multi-path arterial progression with variable phase splits and queue-length estimation

Changze Li

doi:10.1080/21680566.2025.2604695

Multi-band model for multi-path arterial progression with variable phase splits and queue-length estimation

2026-01-22

0

PRE

AI

DOI：10.1080/21680566.2025.2604695

Original

Original request for help

Save

Abstract

En 中文

Although multi-path arterial progression has emerged as a valuable tool for signal coordination, most extant studies generally overlook non-progression paths, mainly relying on fixed phase splits and treating queue lengths as independent of signal plans. To address these issues, we propose a multi-band model for multi-path arterial progression (MBMP). MBMP models queue evolution under varying phase splits, concurrently linking adjacent intersections and preventing spillovers. Further, MBMP decomposes paths into intersection-phase pairs, which act as basic units for local green-band generation. The model utilises effective green bands to capture the queue effects of adjacent plans and overlapping bandwidths to model path connectivity. Non-progressed phase pairs are sequenced to minimise disruption, and a heuristic is developed to enhance efficiency. Subsequent numerical experiments reveal that MBMP consistently mitigates delays and stops compared with MULTIBAND and multi-path band. Sensitivity analysis further confirms the effectiveness of the proposed MBMP, even at low connected-vehicle penetration.

Keywords:

Arterial signal coordination

multi-path progression

queue-length estimation, variable phase splits

Multi-band model for multi-path arterial progression with variable phase splits and queue-length estimation

Abstract

Journal

Organization