Fuzzy

Evaluation of comfort conditions in architectural design assisted by fuzzy logic
Oscar A. Rosas Jaimes∗ , Silverio Hern´ ndez-Moreno† , a ∗ and Alejandro Vilchis G´ mez∗ David J. Delgado Hern´ ndez a o
de Ingenier´a ı de Arquitectura Universidad Aut´ noma del Estado de M´ xico o e Cerro de Coatepec s/n, Toluca, M´ xico e Corresponding author: oscar.rosasjaimes@yahoo.com
† Facultad ∗ FacultadAbstract— We present a fuzzy inference system (FIS) that assists in the evaluation of comfort levels in buildings. From a suggested sustainability design model, based in LEEDTM ’s considerations and recommendations, comfort is analyzed from six different perspectives: temperature, humidity, lighting, acoustic (noise), air (ventilation) and visual relief. These variables are measured through wellknown units, with convenient definitions established, and fuzzified in order to relate them through inference rules to achieve a qualification of the comfort level based on architectonic criteria. Real data from indoors of mexican buildings allowed to evaluate the model’s performance. The result is an expert system capable of ranking either existing or in development buildings. Keywords: Comfort,fuzzy inference system, architectonic design

I. I NTRODUCTION Sustainability can be defined as the ability to meet current needs without compromising those of third and future parties (Herman, 2002; Natural Resources Institute, 1992). Although it is a concept whose need is becoming increasingly more proven, its assessment has many features of subjectivity, a human concept. In this sense, fuzzylogic (Zadeh, 1965) is able to emulate the subjectivity of human thought through establishing a generalization on the membership degree of the elements of a set and on the level of truth of inference-type logical propositions. In general, attempts to evaluate the degree of sustainability achieved by a project or structure are related to the knowledge that an expert may have on the subject. In orderto simulate human reasoning, fuzzy logic also needs the synthesized knowledge of the variables that can be observed on the field of interest, and produce thereby an emulation of the qualification that an expert could give through a process of thought (Zimmerman, 1991). With reference to sustainability, the design and construction of buildings require the use of various method-

ologies and modelsto make them sustainable, i.e. to include criteria for sustainability throughout the entire lifecycle of the building (Hern´ ndez-Moreno, 2010). One of a these methodologies is known as Leadership in Energy and Environmental Design (LEEDTM ), a method used not only to reduce the environmental impacts caused by the buildings but also to improve and assess the conditions of welfare and functionalityof the building (United States Building Council, 2009). It may both save resources of all kinds and improve the quality of life of users and the environment. Basically, this methodology considers six strategies: site, water, energy, indoors comfort, construction materials and scrap management. These ideas are not new (Marsh and Raines, 2009). The modern beginnings on serious works to improve theconditions of life of buildings began around 1993 (CSA, 2001). For example, the international standards organization (ISO) has developed the standard ISO/TC59/SC14 concerning ”life design” to assist in the design of buildings and to improve their conditions of use and performance (Performance Based Building Thematic Network, 2005). However, there is little work where sustainable design and fuzzylogic have been combined. Apart from Roscia (2008), an exhaustive search does not cast more results in this regard. The present paper deals with two main focus to these issues. On one hand, it raises criteria relating to sustainable design for Mexican buildings. Even though we have other methodologies as inspiration (BREEAM, CASBEE, Green Star, DGNB (Delgado-Hern´ ndez et a al., 2010)), the...