Research On The Construction Land Demand In Land Utilization Planning

Land is an important natural resource for human’s living and developing. Theforecast of construction land demand is the core content of urban land use planning.Following the rapidly development of our country’s economic and urbanization, thedemand of construction land is growing quickly. But the predictive accuracy ofconstruction land demand forecast is still at a low level. To improve the forecastaccuracy, we fully considered the concept and features of construction land. We havemade a literature review of construction land demand forecast. We combine thecharacteristic of grey prediction model, and fully considered about the constructionland’s impact factors of economic, society in multiple linear regressions. In order tocover the shortage of lacking of long-term forecast accuracy in grey prediction, wehave forecast the scale of constrction land by the prediction of impact factors. Thisarticle uses Grey-Markov model and multiple linear regression models to forecastthe demand of construction land. We choose Tianjin as the target city to calculate theconstruction land demand.According to the development of economic and society and the status ofconstruction land of Tianjin, we fully consider the influencing factors and have chosenurban population, GDP, fixed asset investment and urbanization as the maininfluencing indicators by CSF. We use the multiple linear regressions method todetermine the linear relationship between the variables, and supplemented byholt-winter model to predict the influencing factor. After that, we use ridge regressionto correct the linear models. Meanwhile, we use gray model to predict the demand ofconstruction land, and supplemented by the Markov chain to improve the predictionresults. According to construction land condition of Tianjin, we combine the twomodels to conclude the demand of construction land of Tianjin from the year2011to2014. After that, we analysis and compare the two results. The results show that themethod which combines multiple linear regressions and Grey-Markov model issimple, easy, and suitable for urban residential land demand forecast.