A Resilience-Based Hybrid Simulation-Optimization Framework for Decision Support in Railway Disruption Management

Railway networks in the UK are highly susceptible to disruptions, with delays and cancellations spreading rapidly across interdependent services due to high network congestion. The growing demand for rail travel has pushed capacity to its limit, reducing operational flexibility and overall resilience. At present, Network Rail (NR) manages disruptions through predefined contingency plans that include alternative timetables for particular scenarios and locations. However, selecting and applying these plans remains a manual process that relies heavily on operator judgment, often resulting in slow responses and operational inefficiencies. To overcome these limitations, this study introduces a hybrid simulation-optimization framework for near-real-time disruption management. The framework integrates a validated delay-propagation simulation model with a genetic algorithm (GA) to identify near-optimal service alteration strategies under disruption. A composite performance indicator is introduced to capture the timedependent evolution of network functionality during disruption, enabling resilience-based evaluation of vulnerability, survivability, and recovery. The framework is applied to a real-world major disruption on the West Coast South Route in the UK and benchmarked against historical operational data. Results show that the optimized strategy achieves a substantial reduction in service cancellations (79.01 %) and in the number of compensationeligible affected services (250 %), while maintaining comparable total delay minutes. Analysis of the resilience curve indicates reduced vulnerability, enhanced survivability, and faster recovery than in the historical case. These findings indicate that the proposed approach can support timely, targeted operational interventions and improve network-level resilience during disruptions. The framework offers a practical decision-support tool for railway operators managing disruptions in highly congested railway networks.

Keywords: Railway Disruption Management, Hybrid Simulation-Optimisation Framework, Genetic Algorithm, Reactionary Delay, Discrete-Event Simulation, Railway Performance Optimisation, Delay Propagation, Resilience.