Decarbonizing Urban Logistics Systems through Closed-Loop Hybrid Simulation: Evidence from Tehran Metropolitan Core

Urban logistics systems in densely regulated metropolitan cores contribute significantly to greenhouse gas emissions, local air pollution, and network delay. This study proposes a leakage-safe closed-loop hybrid simulation framework integrating System Dynamics (SD), Agent-Based Modeling (ABM), and Machine Learning (ML) to evaluate decarbonization pathways in Tehran Metropolitan Core (Tehran Region ۱۲). The framework is executed under a protocol that separates hindcast and projection, avoids clamping endogenous variables, and uses lag-only ML surrogates. Scenario tests (relative to baseline) show that technology levers dominate emissions outcomes: Electrification (E) reduces CO۲ by ۹.۳۸۰% by ۲۰۳۰, while Electrification + Cleaner Electricity (EG) strengthens the reduction to ۱۱.۴۱۷%. Fleet Standards (FS) deliver the largest local air-quality improvement, reducing PM۲.۵ by ۱۷.۲۴۳% on average (۲۰۲۴-۲۰۳۰) and ۳۴.۶۴۴% by ۲۰۳۰. Operational instruments also improve logistics-relevant performance: Access Charging (AC) reduces all-trip network time loss by ۱۵.۸۴۴% on average and ۳۱.۲۷۹% by ۲۰۳۰. All KPIs are computed for total trips; logistics relevance is interpreted via network-level indicators and a conservative freight-like proxy range for Region ۱۲ (LB-۵.۶% and UB=۱۱.۴%). The methodology serves as a compact policy sandbox for sustainable urban logistics decarbonization under fiscal and enforcement constraints.

کلیدواژه ها:

Urban Logistics ، Decarbonization ، Hybrid Simulation ، System Dynamics ، Agent-Based Modeling ، Machine Learning ، Regulated Metropolitan Core