摘要
采用多元回归分析法和BP神经网络建立沉降预测模型,依托泰安抽蓄电站面板堆石坝工程案例进行坝体的沉降预测。通过图表对比及分析可知:两种方法均能实现坝体沉降值预测,但相比而言,神经网络模型因其强大的计算能力及非线性拟合的特性,预测误差值范围仅在0~2 mm,实现了更高的精度预测,而多元回归模型的建立则体现了不同因素相对于沉降值影响的相关性程度,为坝体沉降预测提供了有力的保障。
The settlement prediction model is established by multiple regression analysis method and BP neural network.The settlement prediction of the dam body is carried out based on the case of Taian Pumped Storage Power Station.Through the comparison and analysis of the chart,both methods can predict dam settlement values.However,due to its powerful computing capacity and nonlinear fitting characteristics,the prediction error value of the neural network model is only 0-2 mm,which achieves a higher precision prediction,while the establishment of multiple regression model reflects the correlation degree of different factors relative to the settlement value,providing a strong guarantee for dam settlement warning.
作者
柳丹霞
滕蕊晴
兰昱佳
曹明杰
LIU Danxia;TENG Ruiqing;LAN Yujia;CAO Mingjie(Quzhou City Rural Water Conservancy Management Station,Quzhou 324002,China;Changshan Rural Water Conservancy Management Station,Changshan 324200,China;College of Water Conservancy and Environment Engineering,Zhejiang University of Water Resources and Electric Power,Hangzhou 310018,China)
出处
《浙江水利水电学院学报》
2023年第3期37-42,共6页
Journal of Zhejiang University of Water Resources and Electric Power
关键词
沉降预警
多元回归分析法
BP神经网络
sedimentation warning
multiple regression analysis method
BP neural network
