【摘要】 研究目的:预测城市扩张规模,从而为合理调控城市建设用地扩张提供决策依据。研究方法:通过Granger检验提取城市扩张的驱动因子,利用BP神经网络结合小波处理对建成区扩张趋势进行预测。研究结果:未来几年长沙市新增建设用地数量将呈递增趋势,以年均约27 km2的速度扩张。研究结论:与GM(1,1)模型及回归模型预测方法相比,BP神经网络结合小波处理的预测精度最高,其预测结果可为城市制定相关的政策提供科学依据。更多还原

【Abstract】 The purpose of the paper is to study the trend of urban expansion,thus provide reference for decision-making of regulating urban expansion.Methods of the Granger test and BP neural network and wavelet treatment were employed respectively to collect the driving forces and to forecast the urban expansion.The result indicates that the quantities of the land for urban construction will increase at an average speed of 27km2/y in the coming years.Compared to the GM(1,1) model and the regression forecasting methods,BP neural theory combined with wavelet treatment on data provides the highest precision in the forecasts.As a result,it can provide scientific basis for policy making of urban development.更多还原