A Simulation-Based Decision-Support Framework for Optimizing Bridge–Ferry Operations Under Maritime-Induced Interruptions: The Port Said–Port Fouad Corridor

This study presents a field-informed simulation-based decision-support framework for improving transportation operations within the Port Said–Port Fouad bridge–ferry crossing corridor in Egypt. The investigated corridor represents an interruption-sensitive multimodal transportation system where traffic performance is strongly influenced by maritime navigation activity, bridge-closure events, ferry batch-service operations, fluctuating travel demand, and adaptive traveler behavior. The proposed framework integrates AIS-assisted operational characterization, SUMO-based microscopic traffic simulation, adaptive traveler redistribution, congestion-spillback analysis, XGBoost surrogate modeling, and multi-objective optimization within a unified analytical environment. AIS data were used to identify representative vessel-passage events and bridge-closure periods that supported field calibration of the simulation framework. The methodology explicitly represents bridge-capacity interruptions, ferry operational constraints, multimodal demand redistribution, and corridor-wide congestion dynamics. To reduce the computational burden associated with repeated simulation evaluations, an XGBoost surrogate model was developed to estimate key performance indicators, including transportation delay, vehicle accumulation, ferry waiting time, emissions, and spillback severity. The surrogate model achieved strong predictive performance with a coefficient of determination of R2 = 0.965. Model calibration and within-sample validation were conducted using operational observations collected during a six-day field campaign. The within-sample validation results demonstrated satisfactory agreement between observed and simulated conditions, with an average relative error of approximately 4.8% across major performance indicators. Comparative analyses were performed under existing-operation, rule-based, optimization-based, and adaptive-control scenarios. The results indicate that the proposed framework reduced total transportation delay by 43.8%, peak corridor-wide vehicle accumulation by 65.9%, and estimated CO2 emissions by 17.4% relative to existing operating conditions. In addition, the framework maintained stable performance under increased demand levels and prolonged bridge-interruption scenarios. Overall, the findings demonstrate the potential of simulation-informed decision support and surrogate-assisted optimization for improving operational efficiency, congestion resilience, and environmental sustainability within interruption-sensitive bridge–ferry transportation systems.