Simulation-Based Optimization of Resource Allocation in Seasonal Recreational Facilities Using Discrete Event Simulation and Machine Learning

The study proposes a simulation-based optimization framework to surmount recreational facility operational inefficiencies via spatial design, guest flow, and staff allocation. Adopting Discrete Event Simulation (DES) and Machine Learning (ML), the research optimizes capacity planning and resource allocation in the face of dynamic seasonal demands. A year's worth of operations data was utilized for statistical distribution modeling of visitor interarrival times in RStudio, categorized into low, regular, and high seasons. The simulation model, developed in AnyLogic, uncovered service bottlenecks—particularly at ticketing counters and photo points. Validation results indicated close alignment with real-world operational metrics, ensuring model validity. Actionable suggestions are provided in terms of dynamic employee scheduling and spatial reconfiguration for improved efficiency and visitor experience. By integrating DES and ML, the study contributes to sustainable operations and provides a transferable method for the optimization of service systems in weather-dependent recreational environments.