Deprem sonrası arama kurtarma çalışmalarının matematiksel modellerinde optimizasyon

Turkey is one of the world's most earthquake-prone countries, located on the highly active Alpine-Himalayan seismic belt. In the immediate aftermath of a major earthquake, the efficient allocation and scheduling of search and rescue (SAR) teams is critical for saving lives. This paper gives a brief summary of SAR activities in the aftermath of prior destructive earthquakes that happened in Turkey after 1990 and addresses the "quick search and rescue" problem, focusing on the first few hours following team deployment. We develop a deterministic optimization model for the assignment of professional SAR teams to disaster regions and the scheduling of their operations across collapsed sectors. The main aim of this study is to develop models that account for post-disaster uncertainty in demand and priority. To achieve this goal, we extend our model using minimax regret and alpha-reliable mean excess regret frameworks, enabling risk-aware decision-making under stochastic conditions. To improve the computational efficiency of the alpha-reliable mean excess regret model, we propose a heuristic algorithm that generates high-quality solutions by solving smaller scenario subsets iteratively. The best-performing heuristic solution is then used to warm-start the full stochastic model. Computational experiments are conducted using a simulated earthquake case based on real regional data from Hatay, Türkiye, with demand distributions generated using population-based scaling and randomization. Results show that our deterministic and stochastic models produce interpretable and realistic response strategies, while the heuristic yields near-optimal solutions. Moreover, the warm-start strategy significantly reduces the time required to reach optimality in large-scale instances.