Study On Medium-scale Evaluation On Urban Land Intensive Use Based On ANN

With the rapid development of urbanization and industrialization，the urban landresource in our country faces more and more space pressure and ambient pressure. Therehas been some phenomenans，such as serious expansion of urban outreach，inefficientland use and land use structure is irrational，which caused contradiction between peopleand land，economic and environment conflicts have become increasingly prominent.Therefore，the intensive land use has great significance for the development of theeconomic cycle, building a conservation-oriented society, improving land use efficiency.In terms of evaluation methods of intensive land use，on the basis of the landintensive utilization of some related theories，compared and analysised the characters ofthe model of fuzzy comprehensive evaluation, PRS, TOPSIS, multi-factor comprehensiveevaluation and BP neural network，comprehensive analysis, conducted specifically forthe study of BP neural network model to establish BP neural network model, itssimulation training, the model is applied to come to intensive land use evaluation,whichcan be a good solution to the nonlinear relationship between the various evaluationindexes, eliminate artificial subjective factors to determine the value of the index weightbrings. In terms of divisiing urban functional areas, conducting analysis on thecharacteristics of urban land use in GIS technology as a support, the division method wasstudied on meso-scale urban land functional areas and sample area. From the intensity ofland use, land investment levels, land use efficiency qualitative analysis of variousaspects of the city as a whole, the difference between the various functional areas fordifferent types of functions to select a different evaluation of land can help build acomplete evaluation system.The research selects Tangshan city as an example for empirical study ands selectsthe2009’s Quirk Bird resolution remote sensing image as the data source. In the meso-scale, the central city of Tangshan was divided into three functional areas, residential,industrial and commercial functional areas. And three evaluation index systems wasestablished from land use intensity, land investment and land productivity. Using BPartificial neural network model toestimates the level of intensive land use in differentfunctional areas, and its intensive use of the potential value was calculated. The land usein Tangshan city at a lower level, which has a certain mining space. The intensive degreeof new city is major expansion and transformation of the old city metro. Industrial landintensive use level is higher. The higher intensity area is Technological DevelopmentZone and the eastern industrial zone, mining space is small; commercial intensive landuse level is in the middle. The high intensive degrees is in Tangshan City’s majorshopping malls，there are certain mining space, but relatively small.