<p>Humanitarian supply chains in crisis situations face high uncertainty, infrastructure disruption, and severe time pressure; therefore, decision-making in these environments requires a framework that can simultaneously manage response speed, operational efficiency, performance stability, and network resilience. This research presents an integrated framework based on multi-objective optimization, fuzzy uncertainty modeling, and digital twin to allocate and distribute relief resources in unstable environments in a dynamic, data-driven manner, supporting operational stability. The results showed that the proposed model reduced response time by 30.6% and total cost by 17.1% compared to the baseline, while increasing service level by 25.8% and resilience index by 58.3%. Also, the integration of digital twin resulted in a reduction of response time by 21.6% and demand shortfall by 44.9%. Scenario, Pareto and sensitivity analyses also confirmed the robustness, interpretability and reliability of the model. The findings indicate that the proposed framework can be an effective basis for designing smart, flexible, resilient and sustainable relief networks in crisis situations.</p>