A digital twin-based approach for energy optimisation in medical gas systems in hospital intensive care units
Sustainable Energy Technologies and Assessments, cilt.92, 2026 (SCI-Expanded, Scopus)
- Yayın Türü: Makale / Tam Makale
- Cilt numarası: 92
- Basım Tarihi: 2026
- Doi Numarası: 10.1016/j.seta.2026.105225
- Dergi Adı: Sustainable Energy Technologies and Assessments
- Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Compendex, Geobase, INSPEC
- Anahtar Kelimeler: Intensive care unit, Medical gas systems, Energy efficiency, Digital twin, LSTM
- Sağlık Bilimleri Üniversitesi Adresli: Evet
Özet
Given the increasing energy demands associated with the use of advanced medical technologies in healthcare services, it has become imperative to implement energy optimisation measures in line with global trends. New technologies such as digital twins, system information modelling, the Internet of Things, sensing technologies, and data analytics have become indispensable elements in creating more efficient systems. This study presents analyses using a digital twin-based approach to ensure energy efficiency in medical gas systems through predictive demand control in hospital intensive care units. Within the scope of the study, a digital twin was developed by simulating operational dynamics, taking into account the physical system model, operational dynamics modelling, and control-optimisation modelling. The time series data generated by the modelled digital twin simulation was used to train an LSTM deep learning model that predicts with 89.70% R2 accuracy. The model, implemented through the digital twin simulation, proactively optimises the operating strategy of the medical compressor system using predictions generated by the developed predictive demand control (PDC) algorithm. The use of the generated PDC simulation provides an 18.9% energy efficiency in annual energy consumption compared to conventional methods. Furthermore, a significant improvement was observed in compressor on–off cycles compared to conventional methods. As an advanced approach to system optimisation, the Cycle-Limited Predictive Demand Control (CL-PDC) algorithm was developed by targeting approximately 10 complete start–stop cycles per day under normal operating conditions. This demand cycle resulted in a 13.2% reduction in cycles, 6.0% additional savings, and a +7.8 year extension in contactor life. In the study, the data production process was created by adhering to patient admissions and discharges to the ICU, changes in patient status, and medical gas demand, while also prioritising patient safety. This research highlights the potential of operational digital twins to enhance energy efficiency in hospital infrastructure.