The Spatial Dynamics Of Extended Simulation Study Based On CA Model Of The Central Cities Of Hadaqi Industrial Corridor

The analysis and research of land-use change has become one of the hot issues, thecore of research is how to forecast the tendency of land-use change reasonably andaccurately. Traditional GIS models have great advantages to solve the problem of thespatial relationship,however, its advantage is not obvious for the complex problem suchas land use dynamic change. Cellular automata (CA) model is a "bottom-up" dynamicsimulation model. It can make up the shortage of the GIS in the geographicalphenomena of the complex spatial dynamic changes.The characteristics of CA modelare more appropriate in the use of simulation of land use evolution aspects. CA modelthat integrated with GIS will improve the environment of CA simulation and make theresults of the simulation analysis more accurate.In this paper, the object of study is the central city of Hadaqi Industrial Corridor.The basic source of information is the land-use dynamic database-based. Establishedland-use change simulation model combined with GIS and neural network. Thebeginning data of the model is the land-use spatial distribution in1989. The test data is2010data. We will get the parameters of the model to simulate the initial data and testdata. Finally, predict the central city of Hadaqi Industrial Corridor land-use spatialdistribution in2015.The basis of the data source is Landsat TM images, the central city of HadaqiIndustrial Corridor land-use dynamic database is created through RS and GIS. Thedatabase includes the land-use classification matrixes and the impact factor matrixes in1989,2006and2010. In order to better predict, we get and analyze the land-use maps,at last we get the land-use characteristics of the experimental zone. The experimentalarea of land-use transfer degree is analyzed, then find the direction and magnitude ofland-use change. It laid the foundation for the next modeling and prediction. Theconstruction of the land-use dynamic extension of the model is made by three-tierstructure of the neural network. The three-tier architecture using neural networks toestablish a land-use dynamic expansion model, validated prediction parameters topredict the experimental area of land use change. The accuracy obtained by this methodis relatively high, and the development process on a simulated city and the accuracy of the forecasts of the future development of the city was also higher, so I can play apositive role in urban planning decisions.