Study on the driving forces and prediction of built-up area for urban expansion in Kunming

Based on the analysis of driving forces of urban land expansion by Principal component analysis (PCA), this paper established a predicting model of urban built-up area for future by using socio-economical data. Being good at the performance of nonlinear approximation, artificial neural network (ANN), especially the back propagation algorithm (BP), is applied in the prediction of bulit-up land and had attained satisfactory results. Taking Kunming for example, the results showed that the urbanization is the decisive factor influencing urban land expansion, and a predicting model combined PCA and BP-ANN used to predict urban built-up area in the year of 2009-2015. The method employed in this paper can provide a reference to study on urban land expansion for urban development and planning in the inland cities lacking of multi-sources data.