Study On Dynamic Forecasting Of Urban Space

City is a residential gathering land with a certain size of population which mainly consists of non-agricultural people and is a special form of settlement. No matter how large it is, a city always plays an important role as a focus or core in its sphere of influence. As a result, the development of radiation around the region would be promoted. "Better City, Better Life" has been verified. However, in the urbanization process, particularly in some phases of productivity rapidly improving, a series of urban problems have been raised all over the world, which increasingly attracted many scholars home and abroad. Some new techniques and methods have been used to research various aspects of urban problems.Derived from Study on mining state transition rules and forecasting methods from the long time series of spatial data, one of Natural Science Foundations of Shandong Province, the paper gathered some methods and research which have been used in this field, such as CA, Markov and so on, and then innovated and improved them based on the methods of SDM.In the study of prediction methods on urban land use, there are two respects to research. Firstly, local transformation rules of the cell state space are extracted from long time series sequence map. It is mainly about that spatial data set of state transition rules are automatically extracted from the timing section according to CA. Secondly, the space prediction method based on local state transition rule will be put forward. As the rules of each state transition correspond to a corresponding probability space, that is a extracted random rule, and Monte Carlo simulations will be intended to predict the spatial dynamics calculation respectively.In the study of prediction methods on the geographical expansion of urban space, the main content is the city boundary prediction methods which based on the ideas of spatial diffusion . The spatial diffusion model will be selected according to the timing map data of the historical city border, and than to estimate diffusion parameters to establish the prediction model which could be used to obtain city boundary contour map. Research will be taken into account natural, economic, demographic, planning, transport and many other factors.In the study of prediction methods on the barycenter moving of urban cluster, it is mainly about the predicting method on the trend of moving focus of the future urban agglomeration according to series of statistics data of historical economy and society and map data. A method based on PSO to predict economic barycenter is put forward in this article. Through coordinate segregation of barycenter-moving track of urban cluster, a complex function composed of one trend item and several cycle items is adopted to simulate the change of barycenter's x-, y-coordinates. And PSO is used to estimate the model parameters. It needn't linearize the model in advance.