Forecast And Analysis On Urban Construction Land Scale Of Nanchang City

Land is the material basis of human survival and development, which is the basiccarrier of urbanization and urban construction. China is a developing country whichhas too many people and little land with few vacancies land resources. With the rapiddevelopment of China’s economy, the accelerance of the process of urbanization,urban construction land has been fleetly expanding, as the result, the contradictionbetween the rareness of construction land and urbanization process increasingly acute.The construction land even has become the bottleneck of urban development. It isextremely important to arrange for urban construction scientifically and reasonably. Inorder to implement of the basic national policy of “treasure and rational use of everyinch of the land” and “protect cultivated land”, it is absolutely imperative for us toguarantee food security, maintain social stability and prevent the blind expansion ofthe scale of land used for urban construction to ease the contradiction of populationand land.Currently, most of our cities have drawn up urban land use planning. A newround of general land use planning revision is about to start. The reasonablearrangement of urban construction land index plays an important role in the revision.How to arrange reasonably to provide land for urban construction? It requires a set ofscientific and rational forecasting systems and methods. We start the paper from theactual situation of social and economic development in Nanchang. Considering thefactors that cause changes in the scale of urban construction land are various andcomplex, we screened the relevant factors affecting the scale of urban construction,and then, using the principal components analysis and analysis of correlation methodto determine the main factors: population, construction land area per capita andurbanization rate. In order to improve the rationality and accuracy of prediction, weuse the gray GM(1,1) prediction model, multiple linear regression model and BPneural network forecasting model to predict the construction land in Nanchang Citycombine with the related theories and methods. Besides, the forecast results weremade a comparative analysis, and then is the multiple linear regression model, thegray GM (1,1) model is the worst.The main conclusions are as follow:(1)Thescale of urban construction land in Nanchang expand gradually, as well asthe land use efficiency. However, compared with other mega-cities, its scale remainsrelatively small, and there is room for the improvement in the land use efficiency. (2)Due to the large number of selected factors, there are some linear relationshipsbetween some factors more or less, as the result, it can easily lead to confuse therepeat selection when screening. Therefore, we use a combination of principalcomponent analysis and correlation analysis method to choose the factors. Screenedthe15factors to pick out the major affecting ones, avoided the complex computingand the inaccuracy forecast problems caused by too many factors.(3)Based on the prediction of the scale of urban construction land in Nanchangcity by using the model of grey GM (1,1), it shows that the fitting degree is good. Itcan predict the scale of urban construction land accurately. However, since the greyforecasting model can predict the scale of urban construction land use based on thechanges in time series, ignoring other factors, it makes the grey GM (1,1) projectionof the scale of urban construction land is not suitable for the long-term prediction.(4) In this paper, we establish the BP neural network model to predict the scale ofurban construction land use. The result shows that BP neural network model workswell, function fits well, and has higher accuracy. It suits for the prediction of urbanconstruction land use in the long-term.(5) In this paper, we use multiple linear regression models to predict the scale ofurban construction land in Nanchang, finding that the populations, urbanization rate,Construction land area per capita are the main affecting factors that cause the changesof the scale of urban construction land. By using these factors, we build the multiplelinear regression equations. Empirical study shows that it fit well, has higher accuracy,and computes the result precisely.(6) Compared of the three model, the grey GM(1,1) model, multiple linearregression model, BP neural network model, the result shows that the BP neuralmodel has the highest accuracy, the second is the multiple linear regression model,and the grey GM(1,1) model is the worse. As the fact, we use the BP neural networkmodel and the multiple linear regression model to predict the average scale of urbanconstruction land in Nanchang city in2015and2020.