Optimizing wall insulation material parameters in renovation projects using NSGA-II

Abstract

Renovation works introduce numerous complexities that can only be addressed by those who excel in this specific design task. Such issues as energy consumption which requires examination of excessive alternatives is not of primary concern through the design process due further time limitations. However computational intelligence methods prove to be valuable decision support tools. To this end the current study aims to determine optimum wall insulation material parameters while minimizing optimization targets namely energy consumption and investment costs. To accomplish first energy model of an actual case located in the province of Selçuk was developed using OpenStudio cross platform. Following 54 simulations were run to generate the data base for the given parameters of selected insulation alternatives. Subsequently generated data base was employed to train predictive models of energy generation and investment costs. Finally optimization targets were minimized using NSGA-II algorithm. Results rigorously demonstrate that NSGA-II was able to converge a set of non-dominated set of solutions. © 2017 Elsevier B.V. All rights reserved.