Avlulu ev tasarımı için bir uzman sistem

Abstract

ÖZET Toplumlar tarafından oluşturulan kültürel çevrenin bir elemanı olarak avlulu evleri pek çok yerleşimde yaygın olarak görmekteyiz. Tezin konusu avlulu evlerin sosyo-kültürel faktörlere bağlı olarak tarih içinde ve günümüzde gelişim sürecini irdele mek ve bu tipi bilgisayar destekli tasarım için bir kalıp (pat tern) olarak kullanmak şeklinde belirlenmiştir. Çalışmanın 1. bölümünde konutun oluşum süreci ele alın mış, ikinci bölümde sosyokültürel faktörler başta olmak üzere bina- formunu etkileyen temel parametreler irdelenmiştir. Üçüncü ve dördüncü bölümlerde geleneksel bir konut formu olan avlulu ev tanımı yapılarak tarih içinde çeşitli uygarlık lardan örnekler verilmiştir. Beşinci bölümde geçmişte ve günümüzde kültürel ve iklim sel ihtiyaçlara cevap veren bir konut tipi olarak tüm dünyaya yayılmış olan avlulu evlere çeşitli ülkelerden örnekler veril miştir. Türkiye'de çeşitli illerden (Diyarbakır, îzmir, Kayseri, Konya, ürfa) seçilen avlulu ev örnekleri ise çalışmanın 6. bö lümünü oluşturmaktadır. Yedinci bölüm, günümüzde tasarım ve bilgisayar ilişkisini ele almakta, bilgisayar devriminden hemen sonra gelişen CAD sistemlerinin mimarlıkta kullanımını anlatmaktadır. Sekizinci bölüm, bilgisayarla tasarıma yönelik yapay akıl programlama modellerini tanımlamakta ve kuramsal yapılarını açıklamaktadır. Bu modeller ile ilgili örnekler dokuzuncu bö lümde verilmiştir. Onuncu bölüm, CAD alanında yapılan ileriye yönelik çalış maları ele almaktadır. Onbirinci bölümde ise uzman sistem olarak geliştirilen avlulu ev yapay akıl programı ortaya konulmuş, bu bölümde ay rıntılı olarak açıklanan algoritmaya ilişkin Pascal programlama dilinde kodlanmış program ve bilgisayar çıktıları ekte veril miştir. Onikinci bölümde yapılan kaynak çalışmaları ve geliştiri len algoritma çerçevesinde sonuçlar yorumlanmıştır. - xı -

AN EXPERT SYSTEM FOR DESIGNING COURTYARD HOUSES SUMMARY The environment, which becomes an extension of man's be ing and personality, can be treated as a set of objective topo logical relationships among spaces, objects, event and their attributes. Concepts such as territory, privacy, and personal space all refer to an active use of the environment by people. This can be seen in several house types and COURTYARD HOUSES can be examined as one of them. The first chapter of the thesis is dealed with this so- cio-cultural concept and why it is necessary to generate an ex pert system in order to simulate the design process of these houses. In the second chapter, the housing process is determined, by analysing the mutual and dual impact between man and his en vironment. In this chapter the basic factors which can influen ce the building shape are explained in detail. The cultural, physical, technological, political and economic factors are studied in this context. In the third chapter a detailed description of the courtyard houses is given. The features of these houses demonstrate people's preferences and priorities in the context of their cultural background. By identifying the social relations, characteristics and effect of environment, the basic design principles of these houses are investigated in this chapter. Some examples from the history of Architecture are given in the fourth chapter. First the early settlements in West Asia are mentioned, and cities in Jericho, Khirokitia, Çatalhöyük and Ur are described. Some examples from the past period such as Egypt, Greek and Roman Architecture are also given in this chapter. In the fifth chapter, examples from the Middle East area are given in detail. Also the elements, which correspond the cultural and climatic needs of human beings in early days and nowadays are mentioned in this chapter. The general principles - xii -such as culture, climate-design, and form are the main topics of this chapter. By analysing the examples in China, India, Persia, Kuwait, Malta, Morocco and Saudi Arabia it was possible to identify the main characteristics of these courtyard houses. First of all the cultural structure is defined as the most important mediating variable in determining how things are done and consequently the form of the environment. This relates to values, images, choices and preferences. The second important element is the interaction between climate and design. The ways of sun-protecting, water elements, the compactability and the orientation play important roles in design process. Another principle, which is mentioned in this chapter is `form`. The elements of form such as halls, pedestarian areas, space, gardens and courtyards can be considered as the constituent elements of courtyard houses. The important characteristics of design relate to the interaction between form and environmental factors such as culture, climate and physical aspects of built form such as building materials and technology. This interaction has very complex structure' in nature. If the factors are vigorous then their reflections on form can be observed easily, otherwise careful investigations should be carried out to analyse the nature of the interactions. In the sixth chapter of the thesis a typological approach to courtyard houses is developed, and some examples from Turkey are given. The aspects choosen for these typological analyses are; physical, political, religious and technological factors, which influence the regional architecture. The examples given from different cities of Turkey (Diyarbakır, Izmir, Kayseri, Konya and Urfa) show the differences between courtyard houses. The courtyard houses play an important role in urban Architecture, where the life reflects different values, cultures, life styles and symbol systems of communities. The aim of the thesis is not only to classify the courtyard houses according to some choosen aspects, but to simulate the design process of these houses with the aid of computer. As the computer systems play an important role in design process, it was possible to create an expert system to carry out this simulation process. So, this system can be used to analyse and organize the typology of these houses. - xiii -The Computer Revolution began to emerge in Britain and the united States in the years immediately following the second world war, and gained power in the last two decades. Whereas the Industrial Revolution replaced human muscle power by machines that consume energy, the Computer Revolution is replacing human cognitive power by machines that process information. Architectural CAD (computer aided design) systems are already in quite widespread use, and the introduction of these systems can replace much labor in design process. In the seventh chapter, the impact of Computer Revolution on Architecture is investigated. The philosophy of CAD (Computer aided Architectural Design) systems is explained the relationships beetween architectural design and computer modelling are analysed and the potential roles of computers are discussed at the end of this chapter. The structure of expert systems is the subject of the eight chapter. There are several methods used in these systems such as : top-down or the bottom-up methods. The. examples using these methods or techniques are explained in the nineth chapter. The examples choosen for a further explanation about the above mentioned models are:. A shape independent theory of space allocation. Palladian Grammar. Quenn Anne houses and the shape grammar. Sistematic generation of architectural compositions. The tenth chapter gives information about future of CAD systems, and architectural innovation, which spring from application of techniques available at a particular historical moment to the emerging demands of that moment. The eleventh chapter is dealed with an artificial intelligence programming model for courtyard houses as an expert system. In this model the top-down method in the form of a shape grammar is choosen and the algorithm generated. In this part the constraints that are used to build up the model and the procedure, which the model follows to develope a courtyard house are explained step by step. These steps are: o l.step: Choosing the types. 5 types of courtyard houses have been observed both in Turkey and all over the world. In this thesis the one which has - xiv -L shape has been chosen. Due to the fact that it has almost all the potential problems that might arise in generating courtyard houses. o 2. step : Controling the building site dimensions. The building site dimensions, given by the users are compared with the dimension accepted in the model. The program (GAP) can not be uses unless both dimension are adequate. o 3. step: Choosing one of the L shaped alternative according to the location of street and building. There are four alternatives, which are brought out by rotation :. South-east. South-west. North-west. North-east The user is required to' choose between these alternatives after evaluating the relationships between street and building, in other word the location of the entrance. The entrance should be oriented either directly from the street to the building or from the street to the courtyard and then to the building. o 4. step: Choosing the building's dimensions. There are two alternative ways to choose the dimensions of the building and the courtyard:. The user's choice for the building and courtyard dimensions with the condition that they should be between the given limits.. The program- choose the dimensions from the knowledge base automatically. o 5. step : Space program and clustering of spaces. In this step the spaces that the user needs have to be determined. The clusters among the chosen spaces are organized by the users. The space alternatives and the related symbols, offered to the users are listed below; the entrance: G. the dining room: YO the circulation area: SA. the living room: 00 the kitchen: M. the sleeping room (single) : YOT the bath: B. the sleeping room (double) : YOC the toilett: T. the study room: CO - xv -. the living area: YM Among these different types of activity areas, the user will make his choice according to choosen his/her requirements. An activity area can be chosen several times. Afterwards, these spaces will be clustered in a way, that the groups are limited with 3 spaces. The aim is to connect the spaces in relation with each other. In this study the maximum number of groups for courtyard houses is 4. o 6. step: Determining the dimensions of spaces, the comparisons of group with the length of the building. The minimum widths are compared with the total length, and if the activity areas do not fit the building, the user will be warned, the program automatically turns back to the 5th. step. The user is required to choose new spaces and new clusters. o 7. step: The layout process.. Determining the location priorities of spaces (LPS) in the house layout: These are all determined in the program according to their possible location whether the entrance is through the courtyard or from the building, the program (GAP) will determine the layout. The number of the clusters, generated by the users is an important parameter of the system.. The generation of the first layout according to minimum width of the sapces : The layout process first begins with minimum widths of the spaces. The clusters will be drawn according to their rank in the layout. The lengths and areas are not important in this stage. Each (LPS) will produce an alternative layout. But some layouts should be eliminated if they do not satisfy the given constraints.. The distribution of dummy areas between functional units. If there is dummy areas, these areas should be equally distributed to all of the spaces. Afterwards, the new widths are compared with the limits. The layouts, which do not satisfy the given constraints should be eliminated.. Determining the depth of spaces according to the given criteria. - xvi -There are two alternatives for the depth choice:.. the determination of depth by users... the determination of depth automatically. If the depths are to be determined by the users, they can be calculated by dividing the minimum area value to the other dimensions of the spaces. If those calculated values do not exceed the limits then, they can be accepted as the depths of the spaces. Otherwise there is no solution. If the width is calculated in an automatic process, the program will begin with maximum depth values, given for activity areas. If the depth is greater then the building envelope, the depth line will be accepted as defined by the building envelope, and if it is less then the limits, then the external wall should be modified. In this step if there is no enough space to enter the located spaces at corners, this layout should be abolished.. The refinement of the user ' s choice. After all,. these steps, the user should choose one solution among the produced ones. The doors and walls will be added to the layout. In this study, the entrance is located in the middle the space. The entrances to the spaces will be either from the courtyard or from other spaces. It depends on the given relation. The include files list, subprogrammes, and data files, used in the GAP programme are given in Appendixes. In the 12th chapter, the results are discussed according to the literature survey and in the scape of artificall intellingece. The conclusions are: This technique gives the users the opportunity to cluster the spaces; so it is possible to obtain a building program according to their requirements. There is an interaction between program and user which lets the user to be able to interfere the design process. Thus, the possibility of reaching the better solutions increases. unless the user have special requests, the program automatically create plan arrangements. As a result, the user is able to choose among several - xvii -alternatives, which are created under the defined constraints. The knowledge based computer system show us, the possibility of simulating intellectual activities such as architectural design process and widen the design knowledge in this field. - xviii -