基于径向基函数神经网络方法的城市生态压力预测

针对城市生态压力影响因素复杂,难以对城市未来可持续发展状况做出准确判断的问题,提出了城市生态压力的径向基函数神经网络预测模型,分析了影响城市生态系统的主要因素.以抚顺市1995-2009年数据为基础,验证了模型的准确性并预测了该市2010-2015年城市生态系统的压力情况.研究结果表明:能源消耗指标是影响城市生态系统压力的主要因素;运用径向基函数神经网络模型对训练样本的拟合精度以及对测试样本的仿真精度分别达97.91%和94.16%;抚顺市2015年的人均生态足迹、 生态承载力和生态赤字分别达到7.013、 0.523和6.49 hm²/人.

Urban Ecosystem Pressure Based on Radial Basis Function Neural Network

Considering the difficulty in the estimation of sustainable development due to the complex ecological pressure influencing factors, this paper proposed the radial basis function neural network model for predicting urban ecological system pressure and analyzed the primary influencing factors on urban ecological system pressure. The model was studied on the basis of Fushun’s data during 1995-2009, the accuracy of the radial basis function neural network prediction model is validated and then the situation of urban ecological system pressure was predicted from 2010 to 2015. The results of the study showed that the energy consumption indicator was the primary influencing factor for urban ecological system pressure; the fitting and simulation precision for training and testing samples were 97.91% and 94.16% by using radial basis function neural network model, respectively; the ecological footprint, the ecological carrying capacity and the ecological deficit would be 7.013, 0.523 and 6.49 hm²/cap for Fushun in 2015, respectively.