建设用地预测模型与方法研究  

A Research Into Forecasting Models and Methods of Construction Land

作　　者：曹迎[1] 樊宏[1] 陈钦[1] 

机构地区：[1]成都科地国土资源研究所

出　　处：《国土资源科技管理》2007年第5期81-84,共4页Scientific and Technological Management of Land and Resources

摘　　要：通过对CA(Cellular Automata)模型、回归分析模型、Monte Carlo模型、神经网络模型和灰色模型的对比分析,指出灰色模型的使用限制条件最少,预测准确,是最具有实践意义的预测模型。与单因素预测法相比较,双因素预测法更加全面的反映了各因素对建设用地规模的影响,因而双因素预测法的预测结果更加准确。因此,建设用地空间布局的扩展将成为预测模型的一个发展方向,而权重设置和数理模型参数的选择,则是预测方法主要的发展趋势。With the beginning of a new round of land use planning, summarizing methods and models about construction land forecast is advantageous to advance theoretical research into land use planning and provide theoretical support for the prediction of construction land. Contrastive analysis of CA （Cellular Automata） model, regression analysis model, Monte Carlo model, neural model and gray model demonstrates that gray model is the best for practice for its accurate calculation results and the least restrictive conditions. As for the methods of prediction, compared with the single factor forecast method, the computation result of the dual factor forecast method is better. In conclusion, predicting the extension of construction land on space distribution is one of the important directions for forecasting model. Meanwhile, estimating weight and choosing parameters for mathematic model are very important for the development of forecasting methods.

关 键 词：建设用地 预测 模型 方法 

分 类 号：F301.2[经济管理—产业经济]