采用神经网络模型对天津滨海新区地面沉降预测的研究

在分析天津市滨海新区地面沉降原因的基础上 ,建立了人工神经网络预测模型。模型的训练和检验结果说明 ,该模型的结构合理、精度高 ,可以用于对该区地面沉降趋势的分析预测。预测结果表明 ,在保持现状开采量的情况下 ,塘沽、汉沽和大港 3个区的地面沉降量均有减缓趋势。其中 ,地下水开采量塘沽区应控制在 0 .1 9亿 m3/a;汉沽区则应压缩至 0 .2 8亿 m3/a;大港区即使减少到 0 .2 2亿 m3/a也难以满足 2 0 1 0年规划值 ,因此还应采取综合治理措施

PREDICTION OF LAND SUBSIDENCE IN THE NEW COASTAL REGION OF TIANJIN WITH ARTIFICIAL NEURAL NETWORK MODEL

Land subsidence in Tianjin has restricted the city's development, which attracts many specialists and scholars' attention. Land subsidence in Tianjin has been analyzed and studied by lots of papers and achievements. On the basis of these achievements, in this paper, the author predicts the land subsidence using artificial neural network model, and discusses on quantitative analysis. The results offer some scientific basis and references for the decision making departments.