A comparative study on optimizing multi-generation systems for zero energy buildings in the USA, South Korea, Canada, and England using machine learning and response surface methodology

The study aims to meet building energy demands by implementing diverse renewable energy systems to minimize pollution emissions. A novel multi-energy production system combining Rankine and Kalina organic cycles is proposed to address the energy requirements of three office buildings, a school, and a hospital. Using advanced optimization techniques, including machine learning-based approaches, the building designs are tailored to deliver efficient energy solutions. This research evaluates the system's performance across Liverpool, Vancouver, New York, and Busan, reflecting a range of climatic conditions. The optimized system achieves an exergy efficiency of 31.72 % and an annual cost of $306.33 per hour. It generates 100,740 kWh of electricity, 963,292 kWh of heating, and 123,252 kWh of cooling, effectively meeting energy needs year-round. Additionally, the system's energy storage capacity is analyzed for supplementary applications, offering a comprehensive approach to sustainable energy supply. By integrating renewable sources, optimizing building designs, and accounting for diverse climates, the proposed system demonstrates significant potential for enhancing energy efficiency and sustainability across varied building types.