Measuring Occupants Activities-Generated Carbon Emissions in Healthcare Facilities Using Deep Learning

This article proposes a method to measure occupants’ activities-generated carbon emissions in healthcare facilities using deep learning. The method employs a Restricted Boltzmann Machine (RBM) and a Deep Belief Network (DBN) within the SGAM framework to extract useful features from high-dimensional data and generate predictive and evaluation models. A motivating case in a hospital in Tianjin, China is used to demonstrate the necessity of measuring occupants’ activities, which involves the development of an information integration system with collection, monitoring, operation, and control functionalities. The platform collects data from IoT devices and O&M platforms to form a database, which is updated hourly. The evaluation model is used to determine whether the model needs to be updated. The proposed method provides a way to monitor building operations and control strategies to reduce carbon emissions.