Cities have grown and changed by intricate urban activities, and complex urban land-use patterns have formed. It is reasonable to suppose that complexity is an essential quality of cities. In this paper, the authors try to generate urban land-use patterns using an expanded cellular automata model. The proposed model is composed of three urban activities (residence, business, commerce) which interact economically. Each cell has three kinds of potentials corresponding to urban activities described with real numbers. These potentials can be considered to indicate populations or building areas. Potentials of each cell change by interactions, and each cell takes one of the following three actions: increase, disappearance, and no action. Using this model, urban land-use patterns can be generated. Comparing illustrated results with real cities, it can be said that this proposed model can simulate complex patterns of cities by relatively simple rules.
