Fuzzy model in urban planning JASNA PLEHO, ZIKRIJA AVDAGIC Faculty of Electrical Engineering University Sarajevo Zmaja od Bosne bb, Sarajevo BOSNIA AND HERZEGOVINA Abstract: - This paper presents application of a fuzzy logic in urban planning. Urban environment quality evaluation is an important part of environmental planning and management. Traditional theory does not give as good evaluation as the fuzzy set theory, which provides the basis for urban planning. Most information related to environmental evaluation has the spatial component, which is why GIS is widely used in the evaluation of the environment. Integration of the GIS and fuzzy set theory has been used lately in evaluation of urban environment quality. Fuzzy set theory is used in analysis of the environment because of its ability to manage imprecise, insecure and ambiguous data. Key-Words: - Fuzzy logic, urban planning, GIS, urban evaluation 1 Introduction Evaluation of the urban environment is very important part of urban planning process and it is inevitable part of decision process in urban planning. The main problems of the urban planning could be identified throught evaluation of urban environment. Since conventional evaluation of the environment deals with numerous classification it results in losing the results. Fuzzy logic sheds a new light on urban planning process and evaluation. GIS, geographic information system, has been part of urban planning in the last 15 years. Integration of the GIS and fuzzy logic give the best results in evaluation of the urban environment. Problems in urban planning deal mostly with linguistic concept, for instance, bad urban quality, good urban quality, etc. These terms are fuzzy, vague. Fuzzy logic is used to analyse urban environment because of the ability to manage imprecise, insecure and ambigous terms. This paper describes the use of the fuzzy logic and GIS in urban planning in evaluation of urban quality in Sarajevo Canton, Bosnia and Herzegovina. 2 Problem Formulation Urban planning is the science of managing and directing city growth. Urban, city, or town planning is the discipline of land utilisation planning which explores a very wide range of aspects of the built and social environments of urbanised municipalities and communities. There is more and more population on Earth, less employee needs, less natural resources... Task of the planning process is to organize life, urban life in shortterm two years, medium-term 5 years, and long- term 15 years. Planning is to predict possible number of people in space and fulfill people needs for living, working and other activities, with providing needs for infrastructure (water, energy, traffic, etc). If one draws and colours all that in geodetic map there is no free space left. The main part of urban planning process is evaluation of environment, mostly urban environment. Use of standard methods leads to loss of information and not precise evaluations. Urban planning is a complex proces which consists from different physical, social and technical aspects and is continous evaluation and decision making process. In urban planning you could deal with problems such as finding the best location for city dump. In that problem it is important that the city dump is accessible and at some distance from schools, hospitals, living area. Since terms like accesible, distance are fuzzy criteria the best solution to such problems gives the fuzzy theory. In urban planning evaluation process it is neccessary to indentify key problems which infect urban developement. By identifying the problems we got insight in whole process of urban planning. GIS could be defined as mapping system that uses computers to collect, store, manipulate, analyze, and display data. For example, GIS can show the concentration of a contaminant within a community in relation to points of reference such as streets and homes. GIS has been for the last 15 years a key part of the planning process since most of the information in urban planning deal with space and geographical information. GIS is used in evaluation of the envinronment. With integration of the GIS and fuzzy theory we get the best solution for urban evaluation problems. ISBN: 978-960-6766-57-2 156 ISSN: 1790-5109
3 Problem Solution processing and environment data analysis. In the end the GIS is implemented. In the course of GIS implementation, the structure of raster data are observed, each indicator is an individual layer used for the fuzzy overlay operations. In this process each factor of evaluation is presented as a raster data layer. Each cell of the network is observed as an alternative and each cell value is the result of criteria. The membership value of each cell is calculated on a basis of the membership criteria function. There are two criteria in the evaluation: pollution of environment and social environment. There are sub-criteria, which affect the pollution of environment: air pollution, water pollution, waste disposal and noise. Sub-criteria selected as social environment are: urban population and green area coverage. Figure 1. BiH position in Europe Because of the data availabity the research was carried out in the area of the Sarajevo Canton, Bosnia and Herzegovina (Figure 1.). The first step in urban environment quality evaluation is to identify relevant environmental components (air, water, noise, waste substances, green area) and than to establish relevant evaluation criteria by using these components. The criteria are used for the evaluation of the purpose of the problem decision-making. A valid evaluation is based on the criterion of rational evaluation. The way the evaluation criterion is selected will directly affect the reliability of evaluation results. A series of indicators measuring the quality of urban environment have been developed up to date. Some indicators are measured by the use of human perception of environmental quality. The data used in this paper were taken from analysis which had been carried out in the course of the creation of the Spatial Plan, as well as from relevant institutions which are part of this complex system: the Federation Meteorology Institute, the Cantonal Ministry of Spatial Planning (KEAP Study, Cantonal Plan of Sarajevo Canton Environment Protection), the Sarajevo Canton Development Planning Institute. The paper uses satellite images and ERDAS Software for classification of satellite images in order to obtain the green area coverage. ArcMap 9.1. Software was also used to process and display GIS data. (Figure 2. shows buffer around noise area) Environmental indicators are created in order to measure the environment quality criterion. In general, the selection of indicators is based on following principles: 1. Indicator must reflect main aspects of the urban environment quality 2. Indicator must reflect the human response to the environment quality 3. Data required for indicators must be in such form to enable the researcher to collect them 4. Indicators must be comprehensive both to urban planners and wider population It is necessary to select the membership function, set the weight, materials and data. Then follow the data Figure 2. Noise buffer in Sarajevo Canton ISBN: 978-960-6766-57-2 157 ISSN: 1790-5109
MATLAB- Fuzzy Toolbox Software was used for the analysis of final results. In the course of the research a survey was carried out among the participants in the process of urban planning (15 respondents). Basic factors that affect the urban environment were set and the survey included the following factors: air pollution, water pollution, noise, and green area coverage, percentage of solid waste disposal and population density. The survey provided criteria values, which have an impact on the decision-making process in evaluation of the urban environment quality. AHP method (Analytical Hierarchy Process Method), (Saaty(1980)), which is based on three principles: decomposition, comparative judgement and priority synthesis, was used. This is a mathematical method used for selection of priorities among different alternative decisions through paring up of comparative decision elements with preservation of common criteria. Figure 3. Fis Editor for environment evaluation, Mamdani Type Membership function editor (sets title, operating volume, display range and parametres of each input and output, as well as function which describes each input and output). Following the above mentioned processes, there were data analysis and evaluation of fuzzy values to obtain a final crisp value, the environment quality which as such is then implemented in any GIS Sofware and any well laid out map which gives a clear result not only to urban planners but also to non-specialists. There are 5 steps in the fuzzy logic: 1. Phase up of entry variables 2. Application of fuzzy operators 3. Application of fuzzy method 4. Aggregation of all outputs 5. De-phase up In the MATLAB package, the process is carried out in five modules and it is applied through the search of favourable urban environment. Two factors were taken as examples (environment pollution and social environment). Following items provide Matlab examples. Figure 4. Membership function for Environment variable, Gauss membership function Rule editor (form which provides input of if-then rule by selecting membership function for each variable and operation in use, result which follows in given case with capability to rank results by weight input) FIS editor (inputs and outputs are created in this form, operation methods and/or implication, aggregation and de-phase up methods are set). ISBN: 978-960-6766-57-2 158 ISSN: 1790-5109
Figure 5. Rule Editor for environment evaluation Figure 7. Surface viewer, viewer of environment evaluation area Rule results viewer (graphic display which shows how set operations over functions based on input values again result in de-phase up value). The figure 8. shows FIS file, which serves to phase up two parameters. Selected membership function is gaussmf. Gaussmf is Gauss curve with following syntax: y = gaussmf (x, [sig c]) Symmetric gauss function depends on two parameters σ and c, and it is given in following equation (1): f (x; σ, c) = e 2 ( x c) 2 2σ (1) Figure 6. Rule Viewer for environment evaluation For example: x=0:0,1:10; y=gaussmf(x,[2 5]); plot(x,y) xlabel('gaussmf, P=[2 5]') Diagram Viewer (three-dimensional diagram of dependence of one output in function of two selected inputs). Figure 8. Gauss curve ISBN: 978-960-6766-57-2 159 ISSN: 1790-5109
Result of using fuzzy results and GIS gives next map that shows best urban quality contidion References: [1] Z. Avdagic, Artificial Intelligence & Fuzzy-Neuro- Genetic, Grafoart, 2003 [2] Wuhan- Dou Kaili, Fuzzy evaluation of urban environmental quality Case study Wuchang, Septembar 2003. [3] Institute for Canton Sarajevo planning, Spatial plan for Sarajevo Canton from 2003. 2023, Sarajevo 2006. [4] Longley, P., Goodchild, M., Maguire, D. and Rhind, D., (2002), Geographic Information Systems and Science, John Wiley&Sons, Ltd. England Figure 9. Evaluation of urban quality Sarajevo Canton 4 Conclusion Better planning of urban environment is based on a complete and precise understanding of environment quality conditions and urban area. Most researches focus only on some aspects of urban environment quality, for example, urban traffic environment, urban population environment and urban community environment. Thanks to fuzzy set theory an ideal core for solving the fuzzy processes in the course of environment quality evaluation is secured. However, major application of urban environment quality evaluation is focused only on quantity perception and not on spatial perception of urban environment quality. The application of GIS and fuzzy set theory in environment quality evaluation is still scarce. The application of new methods, such as fuzzy set theory, would give better and comprehensive results which would provide for more possibilities and space in processes which are part of cantons, towns and local communities development planning. ISBN: 978-960-6766-57-2 160 ISSN: 1790-5109