柴河水库大坝坝基渗流监测数据研究

柴河水库大坝右坝段坝基存在渗漏问题,为了水库大坝的安全建设和运行对大坝监测数据进行研究非常重要。采用逐步回归分析法对大坝坝基测压管多年监测数据建立了统计回归分析模型,计算结果表明,观测值与拟合值统计复相关系数较大,估计标准误差较小,模型有效地反映了坝基渗流的变化规律和发展趋势。为了监控大坝的安全运行和辅助决策,利用小波神经网络建立了有效的坝基渗流量预测模型,计算结果表明,该预测模型收敛速度较快、预测精度较高,能正确地模拟和预测大坝的渗流量。

Study on Monitoring Data of Seepage Quantity of Dam Foundation in Chaihe Reservoir

There is a leakage question in the right dam foundation of Chaihe Reservoir,so the study on the monitoring data is very important to the safety building and running of the reservoir dam.A statistical regression analysis model for some piezometric observation data of the dam foundation is established by the stepwise regression analysis method.The computation results indicate that the statistical multiple correlation coefficients between the statistical regression data and measured data are larger,and the estimation standard errors are smaller,at the same time,the statistical regression analysis model could efficiently reflect the changing rule and developing trend of seepage in dam foundation.In order to monitor the safety running and assistant decision-making of the dam,the wavelet neural network is used to build an effective predication model of seepage quantity.The computation results show that this model is faster in convergence speed and more accurate in prediction than BP neural network,and could simulate and predict the seepage quantity correctly.