ENUMERATING A DIVERSE SET OF BUILDING DESIGNS USING DISCRETE OPTIMIZATION
Elaine T. Hale, Nicholas L. Long
Numerical optimization is a powerful method for iden- tifying energy-efficient building designs. Automating the search process facilitates the evaluation of many more op- tions than is possible with one-off parametric simulation runs. However, input data uncertainties and qualitative as- pects of building design work against standard optimiza- tion formulations that return a single, so-called optimal design. This paper presents a method for harnessing a discrete optimization algorithm to obtain significantly different, economically viable building designs that satisfy an en- ergy efficiency goal. The method is demonstrated using NREL’s first-generation building analysis platform, Opt- E-Plus, and two example problems. We discuss the infor- mation content of the results, and the computational effort required by the algorithm.
- There are currently no refbacks.