Implementation of Heuristic Algorithms for Simulating Crisis Situations in the Medical System

Abstract

Healthcare systems face significant challenges during crises such as pandemics or mass-casualty events, where resource shortages and patient overflow require rapid, optimized decisions. This paper proposes a simulation-based approach to model and improve hospital resource allocation using a heuristic method—Genetic Algorithms (GA). The aim is to explore how intelligent algorithms can support decision-making under pressure by assigning patients to limited ICU beds and available doctors, considering constraints such as treatment duration, medical priority, and resource availability.The core method involves evolving allocation strategies over multiple generations, using tournament selection, crossover, mutation, and fitness-based evaluation to optimize both resource usage and patient coverage. The simulation is implemented as a desktop application in C#, with a SQL Server database and an interactive GUI that allows users to run scenarios, configure parameters, and visualize outcomes.Compared to a First-Come-First-Served (FCFS) baseline, the GA consistently achieves higher efficiency, treating more high-priority patients and reducing resource bottlenecks. The original contribution lies in the dual-resource optimization model and its integration into a flexible, user-friendly tool. Results demonstrate that heuristic-driven simulations can support better planning and training in emergency healthcare environments.