摘要
针对大坝位移建模预测中存在多重共线性和高度非线性问题,基于小波消噪技术消除噪声对建模的影响,并以某大坝为例,采用PLS-SVM大坝位移预测组合模型进行建模预测。结果表明,该法预测模型效果优于单一预测模型。
According to the problem of multicollinearity and non-linear relationship in prediction model of dam displacement, a wavelet de-noising technique is proposed to eliminate the influence of noise on model building. The combination prediction model of dam displacement wish PLS-SVM is proposed. The result shows that the effect of combination forecasting model is superior to that of single prediction model.
出处
《水电能源科学》
北大核心
2010年第4期81-82,58,共3页
Water Resources and Power
基金
国家自然科学基金资助项目(50879024)
国家自然科学基金资助重点项目(50809025
50539110)
国家科技支撑计划课题基金资助项目(2008BAB29B06)
江苏省"333高层次人才培养工程"基金资助项目(2017-B08037)
关键词
大坝监测
小波消噪
偏最小二乘回归
支持向量机
组合模型
dam monitoring
wavelet de-noising
partial least squares regression
support vector machine
combination model
