Modeling of Power Management Systems for Healtcare Facilities Using Hybrid Fuzzy-Particle Swarm Optimization: A Case Study of Ulin Hospital

Healthcare facilities are known for their intricate energy systems and their significant energy consumption. Therefore, they serve as key areas for achieving the energy conservation goals in the building sector. Accurately forecasting hospital energy consumption loads is critical for successful energy management in healthcare buildings. This paper presents a Modeling Power Management System to numerically model the electrical consumption of Healthcare Facilities (such as Ulin General Hospital, a part of the South Kalimantan Government's Indonesian hospital network), alongside the application of such models to Power Management Systems. This study employs a Fuzzy Logic technique for forecasting electrical energy consumption and cost estimation, specifically applied to hospitals. The fuzzy-based-system, as a form of Artificial Intelligence, is utilized to analyze data collected by Power Management Systems. This system can be employed to predict power usage, for example, where one of the inputs used as a parameter aligns with hospital characteristics such as patient visit data, temperature, and daily electricity consumption. The fuzzy logic techniques employed include Fuzzy Type-1, FuzzyParticle Swarm Optimization (PSO), and Fuzzy-Sealed Annealing (SA) for electrical energy consumption forecasting. This approach demonstrates significant improvements over traditional methods and provide a potential model for broader applications in healthcare energy management. These results suggest that the fuzzy-PSO method is more accurate in forecasting power consumption at Ulin Hospital, Indonesia, compared to the other evaluated methods. The smallest Mean Absolute Percentage Error (MAPE) value using the fuzzy-PSO method for Peak Load Time (WBP) is 10.97%, the MAPE value with fuzzy-SA is 17.51%, and the MAPE value with fuzzy Type1 is 21.75%.