摘要
混凝土坝变形预测模型通常采用谐波函数模拟温度效应,但没有考虑不同年份坝体温度变化对混凝土坝变形的影响。针对此问题,利用有限元模型(FEM)计算水压分量,利用长期气温序列代替谐波函数计算温度分量。考虑混凝土坝变形过程的非线性特征,引入麻雀搜索算法(SSA)对随机森林回归(RFR)模型的参数进行优化,获取最优参数对混合模型进行训练,最终构建一种基于长期气温进行温度效应模拟的混凝土坝变形预测混合模型。结果表明,相较于RFR变形预测模型和基于统计学回归方法的变形预测模型,采用SSA优化RFR的变形预测混合模型具有更好的拟合效果和预测能力。
The temperature effect is usually simulated by harmonic function in the concrete dam deformation prediction model,but it does not consider the impact of dam body temperature changes in different years on the deformation of concrete dams.In response to this issue,the Finite Element Model(FEM)was used to calculate the water pressure component and harmonic sinusoidal functions were replaced by long⁃term air temperature to simulate the temperature effect on the dam response.Considering the nonlinear characteristics of concrete dam deformation process,Sparrow Search Algorithm(SSA)was introduced to optimize the parameters of Random Forest Regression(RFR)model,obtained the optimal parameters to train the mixture model,and finally built a concrete dam deformation prediction mixture model based on long⁃term temperature effect simulation.The results show that the RFR deformation prediction model optimized by SSA has a better fitting effect and prediction ability than that of the RFR deformation prediction hybrid model and the deformation prediction model based on the statistical regression method.
作者
陈旭东
侯阵阵
郭进军
CHEN Xudong;HOU Zhenzhen;GUO Jinjun(Yellow River Laboratory,Zhengzhou University,Zhengzhou 450001,China;Nanjing Hydraulic Research Institute,Nanjing 210017,China;National Dam Safety Engineering Technology Research Center,Wuhan 430010,China)
出处
《人民黄河》
CAS
北大核心
2023年第9期141-146,共6页
Yellow River
基金
国家自然科学基金资助项目(U2040224)
国家大坝安全工程技术研究中心开放基金项目(CX2022B05)
“一带一路”水与可持续发展基金项目(2021nkms06)。
关键词
混凝土坝
变形
混合模型
温度效应
麻雀搜索算法
随机森林回归
concrete dam
deformation
hybrid model
temperature effect
Sparrow Search Algorithm
Random Forest Regression
