An Evolutionary Multi-Objective Optimization Framework for Enhancing Energy Efficiency in Early Phases of Architectural Design Process

Abstract

To achieve high-performance building designs that consider the competitive objectives of energy efficiency, architects must consider optimal use of passive environmental design strategies in early design phases. During these phases, building design decisions are made, and the greatest potential to achieve high-performance building designs exists. Hence, this thesis introduces for architects an integrated evolutionary multi-objective optimization (EMO) framework connecting parametric modeling (PM), building performance simulation (BPS) and building performance optimization (BPO) workflows seamlessly; to automate building design form generation, performance simulation and, optimization processes in conceptual design stage. The applicability and the effectiveness of the developed framework was quantified through a baseline case study model of a large hotel building (570 rooms) located in Cairo, Egypt. The capabilities and limitations were explored through improving five common correlated and conflicted concept and performance objectives of energy efficiency. The objectives were: constraining footprint area (A) to a certain range of area, view quality index (VQI) maximization, total incident solar radiation (I) minimization, spatial useful daylight illuminance (sUDI) maximization, and energy use intensity (EUI) minimization. Additionally, the framework can provide real-time feedback, find the objectives’ trade-off solutions (Pareto-optimal solutions), and visualize the resulted optimal design alternatives (phenotypes) in a proper way for further decision-making process by the architect. With consideration of performance metrics and comparing to the baseline case study pre-measured objectives, the results showed that applying the framework to the case study can constrain A to a certain range defined by architect and accepted by laws, deliver simultaneous improvement in both: VQI ranges from 2% to 25%, and sUDI 0.4% and 23%, while providing reduction in both: I by a percentage of 11%, and EUI ranges from a 3% to 23%, depending on the local site and climatic conditions. Hence, the EMO framework and its integrated parametric simulation-based design workflow is fundamentally changing architectural design process into a faster, performance-aware and more flexible process. It can ease the production of multiple high-performance design alternatives, facilitate the decision-making process, and open up new design scenarios that were inaccessible previously in the early design stages.