A Pareto Based Genetic Algorithm Model for Sustainable Site Layout Design of Social Housing:  SSPM

Yazg� Badem Aksoy¹, G�len �a�da�¹, �zg�n Balaban²
¹�stanbul Technical University, Architectural Design Computing, �stanbul, Turkey
²Singapore University Of Technology And Design, Architecture And Sustainable Design, Singapore

Sustainable architectural design process is more complicated than traditional building design because it has its own regulations that require calculations and comparisons that also need to be considered. For this article, sustainability parameters were accepted as the standards of green building certification systems (LEED and BREEAM), local climate conditions, and local building regulations. There are conflicting criteria according to those sustainability parameters. As a result, Pareto genetic algorithm was chosen to address multiple objectives by incorporating the Pareto concept of domination. An efficient, sustainable design should be developed during conceptual stage, when the most important decisions are made, so that a decision support system can be developed with the help of Pareto-based non-dominated genetic algorithm (NSGA-II) to evaluate several possibly conflicting objectives. This article discusses a model that aims to produce site layout alternatives according to sustainability criteria for social housing units.

Keywords: LEED-BREEAM, multi-objective genetic algorithm, sustainable site layout design.

S�rd�r�lebilir Toplu Konut Yerle�mesi Tasar�m� ��in Pareto Genetik Algoritmaya Dayal� Bir Model �nerisi: SSPM

Yazg� Badem Aksoy¹, G�len �a�da�¹, �zg�n Balaban²
¹�stanbul Teknik �niversitesi, Mimari Tasar�mda Bili�im, �stanbul, T�rkiye
²Singapur Teknoloji Ve Tasar�m �niversitesi, Mimarl�k Ve S�rd�r�lebilir Tasar�m, Singapur

S�rd�r�lebilir bina tasar�m� geleneksel bina tasar�m�na g�re daha karma��k bir s�rece sahiptir. Bunun nedeni birlikte al�nmas� gereken fakat birbiriyle �eli�en pek �ok tasar�m karar�n�n erken tasar�m evresinde al�nmas� gereklili�idir. S�rd�r�lebilir tasar�m�n erken evresinde tasar�mc�ya yard�mc� bir karar destek sistemine ihtiya� ka��n�lmazd�r fakat tasar�m end�strisinde, bilgisayar tabanl� ara�lar tasar�m s�recinin �nemli bir par�as�n� olu�turmalar�na ra�men, erken tasar�m evresi bilgisayar deste�inin en az kullan�ld�� evre olmu�tur. Bu �al��mada, pareto temelli bast�r�lmam�� s�n�fland�rmal� genetik algoritma (NSGA-II) kullan�larak pek �ok tasar�m kriterini g�z �n�nde bulundurmas� gereken tasar�mc�ya karar verme s�recinde yard�mc� olacak bir model geli�tirilmesi ama�lanm��t�r. Bu konuda yap�lm�� di�er modellerden farkl� olarak, LEED ve BREAM arazi yerle�im kriterlerinin yan�nda yerel y�netmelikleri ve yerel iklim �artlar�n� da g�z �n�nde bulunduran modelin, �retilen alternatifleri uygunluk puanlar�na g�re birbirleriyle k�yaslayarak pareto optimal sonu�lar �retmesi, tasar�mc�y� s�rd�r�lebilir arazi kullan�m ��z�mlerine g�t�recektir. S�rd�r�lebilir Arazi Yerle�me Planlama Modeli, karma��k problemlerin ��z�m�nde tercih edilen evrimsel algoritmalar�n s�rd�r�lebilir arazi yerle�iminde kullan�lmas�yla, hesaplamal� tasar�m ve s�rd�r�lebilir mimarl�k kavramlar�n�n arakesitinde yer alan ve her iki disiplinin olanaklar�n� bir arada kullanan, erken tasar�m evresinde toplu konut bloklar�n�n araziye yerle�iminde mimar� y�nlendirebilecek bir karar destek sistemi olarak �nerilmi�tir.

Anahtar Kelimeler: �ok ama�l� pareto genetik algoritma, LEED-BREEAM, s�rd�r�lebilir arazi yerle�im plan�.

Yazg� Badem Aksoy, G�len �a�da�, �zg�n Balaban. A Pareto Based Genetic Algorithm Model for Sustainable Site Layout Design of Social Housing: SSPM. Megaron. 2016; 11(2): 241-253

Corresponding Author: Yazg� Badem Aksoy, T�rkiye

TOOLS

Full Text PDF

Download citation

RIS

EndNote

BibTex

Medlars

Procite

Reference Manager

Share with email

Send email to author

Similar articles

PubMed

Google Scholar