A novel hybrid model to evaluate the location of net-zero energy consumption building based on remote sensing, analysis hierarchical process and machine learning

The design and architecture of buildings are critical stages in construction, with significant emphasis on environmental sustainability and energy efficiency. Achieving a net-zero energy architectural design necessitates careful consideration of the building's location. This simulation-based study introduces a novel hybrid model that integrates remote sensing (RS) techniques with the analytical hierarchy process (AHP) and a machine learning (ML) algorithm (M5 model tree) to optimize and validate site selection for net-zero energy buildings (NZEBs) in Xi'an, China. The proposed approach enhances decision-making accuracy by leveraging RS data and ML-based validation for robustness. Simulated datasets, derived from RS and applied scenarios, were used to evaluate key factors such as solar energy radiation, climatic conditions, green space distribution, accessibility, population density, and proximity to renewable energy sources. The results demonstrated that solar energy radiation, accessibility, and proximity to renewable energy sources were the most influential factors, with weights of 0.40, 0.25, and 0.20 in the AHP analysis and 0.42, 0.25, and 0.18 in the M5 model tree validation, respectively. Thirteen urban areas of Xi'an were analyzed, and the results indicated that Zones 3, 6, and 8 received the highest suitability scores, while Zones 9, 12, and 13 were deemed the least suitable for NZEB projects. The hybrid model demonstrated high accuracy, with a Pearson correlation coefficient of 0.91 confirming strong agreement between the AHP-derived weights and the M5 model tree predictions, validating the consistency and robustness of the decision-making framework. The findings provide practical aspect for urban planners and policymakers, offering a data-driven framework that ensures NZEB developments align with energy efficiency goals and environmental policies.