A Study Of Urban Land Grading Based On The Method Of Fuzzy Clustering

Land is the basic material for mankind's survival, as well as the place on which people engaged in producing and daily life activities. Facing the world today, the issues such as population, resources, environment and socio-economic development all related to land. Rational land use and management have been concerned by all the nations of the world, and has become an important research topic in the scientific community. This will also play a very important role in facilitating the coordinating development of population, resources and environment.As deeper as the reform on land use system in our country, scientific and effective management and use of land has become an urgent task. Currently, various forms of compensating use such as land sale, transfer, lease, tender, auction are being the main form for the use of urban land. The dividing of the urban land hierarchy need to base on their inherent differences in the factors, use appropriate theory, scientific method and means, in order to assess the land use quality, efficiency and spatial distribution intrinsic or the merits of the land. Qualify the difference of urban land economic value. Thus, not only can do the work of compensation use reasonably from the point of full playing to urban land values, provide the scientific basis for the land sale, transfer, fees, tax collection and so on, but also can define the structure and layout more reasonable for urban land development. It is of great significance for improving and deepening the reform of the land use system, improving urban land use planning and urban construction, improving economic, social and ecological benefits of urban land-use.Fuzzy clustering analysis is a kind of methods from the viewpoint of things to explore the classification of number. It classifies the objective things through the establishment of similar fuzzy, fuzzy similarity matrix formation. Traditional clustering analysis is a hard division, pending identification of each object of strict division of a certain category with nature of one or the other, so the type limits of the classification is obvious. In fact the majority of recipients were not strictly in nature, their behavior in class and there are intermediary, suitable for soft division. Based on the fuzzy set theory of fuzzy clustering analysis of this kind of soft partitioning provides a powerful analytical tool. Fuzzy clustering is due to the samples belonged to various categories of uncertainty, the sample expressed an intermediary category, that is, the uncertain description is established for the samples to types, reflected the real world more objectively.Based on the outcome of other researchers, depending on the principle, procedures and methods of land grading established by Ministry of Land and Resources in 2001 called "urban land classification order", on the basis of detailed investigation of natural environment, social status and economic development in Wanzhou District of Chongqing, combining with traditional land grading theory, this research systematically studied the three ways which has been widely applied in describing the process of urban land classification—multi-factor assessment, Gaps premium income and zoning classification method. Take the land grading in Wanzhou district towns as an example, it discussed the rating factors in the selection and role of scores, land grading unit and element of the total score, and paid more attention on the fuzzy clustering analysis in land-level architectural applications. Main content and conclusions are as follow:1. According to site inquiries of experts, define Urban Land grading factors factor in Wanzhou District, after conducting two points Delphi get weights of the Urban Land grading factors and factor in Wanzhou District. On the vertical and anonymous feedback conditions, the improved Delphi scoring method make the opinions of experts convergence more quickly, reducing the scoring rounds, avoiding the phenomenon of score closer to the average consciously by the number of points. Taking the multi-factor comprehensive assessment as the theoretical basis, calculate the effect of factors and factor scores severally. Divide urban land classification modules by "overlay", and calculate the modules total scores.2. Analyses the total score by the method of Modules fuzzy clustering analysis, and define the level of urban land. Firstly standard unit total score by the method of extreme standardization. Secondly build fuzzy similar structure relationship, and transform it into equivalent. Then get the Urban Land Classification results by different values interception fuzzy equivalent matrix. Point out the inadequate of land-level conventional method—total core axis method, frequency and total profile law method. Detail fuzzy definition of statistical methods in the exposition of the fuzzy clustering method. Compare the applicability of the distance definition way and the correlation coefficient definition, so as the coverage of various distance definition.3. Analysis the results of demarcation of land-level by variance methods. Check results with the whole different level of the land division by the method of variance test, when K get the maximum, the plan is best for the optimal partition.4. Combing with Fuzzy clustering methods and variance analysis, divide urban land-level chiefly by quantitative analysis, secondly by qualitative analysis, which change the subjective judgment model in land-level partition process in the past, increase objective reliability of the land-level Classification, and provide a new way for the land-level Classification of Urban land.