Abstract
In model building for urban enviroment, we often enconunter the difficulty of structure identification and the lack of homogeneity in data. To cope with these problems, we are developing a modeling support system to utilize objective information and subjective knowledge effectively. In this paper we propose a fuzzy modeling technique with a visual and stepwise clustering method, and a fuzzy simulation technique for reasonable scenario input and interpretation of the model behavior. We present an application of the proposed techniques to an urban environmental problem.
