Using the Road Constraint Cellular Automata to Simulate the Urbanization in Zhengzhou

Urbanization has resulted in the extensive land cover changes. It is helpful to find out and investigate the discipline of the land cover changes. The study will provide a useful support for the decision making of the regional development. Based on remote sensing images in 2000 and 2007, using RS and GIS, the change of the city from 2000 to 2007 have been analyzed and the model of cellular automata based on the artificial neural network was adopted to simulate the city expansion in this paper in Zhengzhou. In the study the factors of speedway, railroad and the city outer highway were involved. The relation of the road and the changes was analyzed. Apparently the two elements were correlative. Considered the transit-oriented land use development, the road influence factor was strengthened from point to area, according to the fact of Zhengzhou the width of the road buffer was decided as 5 km, 10 km, and 15 km. The buffer maps and the map of the land use in 2000 were transformed into grid, then the sample points were obtained in the GIS and based on the MATLAB, the samples were trained. By the adjustment of the network structure, the precision of the training was reached. After this, all the grid maps were input to the trained model to simulate the map in 2007 and in 2014. The simulated map in 2007 was compared to the factual. The comparison indicated the precision of involving the road constrict is higher than that without condition; the city space would breakthrough the limit of the outer ring and sprawl all around. But the rural resident land would develop rapidly to the southeast and the northwest.