我国耕地利用压力区域差异的RBF模型判定

Analysis on Regional Differentiation of Cultivated Land Use Pressure in China by Radial Basis Function Networks

在对目前耕地利用压力程度度量方法进行分析的基础上,运用人工神经网络(ANN)的理论和方法,构建了ANN模型分析中应用最为广泛的RBF网络,并对2004年中国31个省、市(自治区)的耕地利用压力程度进行了评价。网络运行结果表明,中国耕地利用压力程度的区域差异显著,耕地利用压力程度较高的省(市、区)主要分布在东部沿海地区,耕地利用压力程度较低的省(市、区)主要分布在中部和西部地区。我国耕地利用压力的区域差异主要表现为东部和中西部及沿海和内地的差异。在这些分析的基础上,作者提出了耕地分区保护的对策和措施。RBF模型的应用表明,人工神经网络用于评价耕地利用压力区域差异简便、实用,且避免了人工确定指标权重的主观性,是一条具有发展和应用前景的途径。

Radial basis function （RBF） neutral network is a widely used artificial neutral network model with the feature of self training and strong capability in solving nonlinear problems. On the basis of cultivated land use pressure in China , application of RBF theory and method , adopting pressure evaluating level , the authors develop a neutral network model , in which Chinese cultivated land use pressure level in 2004 is evaluated. The run result of RBF shows that the regional differentiation of cultivated land use pressure in China is obvious , which accords with reality. The first is Beijing with the highest cultivated land use pressure of 1.4237, while the lowest is Gansu with the evaluation of 0. 6121. According to the cultivated land use pressure evaluation, the province （cities）or autonomous regions with the high evaluation value are mainly distributed in the eastern coastal area and the provinces （cities）or autonomous regions with lower evaluation value are mainly distributed in the middle and western areas. Cultivated land pressure status estimated on RBF model can explore distinctively the reasons and the results of cultivated land use press!are changes, which will help administrators to adopt suitable land policies and management measures to alleviate cultivated land protection pressure and improve cultivated land quality. The evaluation results also indicate that application of RBF neural network to assessing regional differentiations of cultivated land use pressure level without assuming parametric relationship is convenient, precise and feasible, which can be an alternative approach to assessing regional differentiations of cultivated land use pressure level.