Probabilistic models for energy efficient building: a case-based support system for schematic design phase

This paper reports on a research aimed at developing a design support system that allows the simple exploration of design alternatives in buildings’ conception phase and the execution of building analysis in terms of effective energetic performance. The system’s core consists in Bayesian probabilistic models which manage a case-based database. Cases express design parameters and energetic performance knowledge and are derived from common simulation software. A prototype Bayesian model has been developed for evaluating the effectiveness of the support system hypothesized in common schematic design scenarios.