Space Simulation Studies, Based On Cas Theory Of Urban Land Use

Understanding and realizing the basal disciplinarian of land use change in urbanization in China and modeling the dynamic process of it have theoretic and practical value to provide decision support for regional development by description and simulation and optimization of dynamic behaviors in process of changing in urban land use.Land use change has its systematic characteristic. Among which complex adaptive character is more important because of activities of human. From the view of Complex Adaptive System (CAS), it could be easier to understand "advanced biologic phenomena" of natural system which is filled with changes, competitions and evolutions etc. Therefore it's more feasible to simulate complex real world.Simulation research is divided into two parts, i.e., dynamic simulation and static optimization. The dynamic part focuses on simulation and prediction of urban growth. From the key idea of CAS, the paper puts forward that the growth, evolution and decay of earth is the result of adaptation and reconstruction to environment. By applying the idea to the real urban growth simulation and based on the Cellular Automaton model, an urban growth model is developed based on adaptability and macroscopical control mechanism. The static part emphasizes optimization to spatial structure of urban land use. The paper puts forward a model for optimizing land use based on genetic algorithm. The model codes the land use image with two-dimension matrix and puts forward the multi-object function that considering function harmony index, proportion harmony index, population index and traffic index of land use.Based on the two models mentioned above, a simulation platform for land use is developed. Visual C++6.0 and MapObject2.2 are employed. The platform is applied to simulating and predicting Xi'an urban growth. The results suggest that the adaptability based cellular automaton model can represent the characteristics and rules of city development, especially for the "Unurbanization" that most research never concerned. Some of them have directive value to urban planning. The genetic algorithm based urban land use optimization model can overcome the subjectivity of layout of land use. The optimized scheme can reach diversified requirements for urban planning.