基于灰色神经网络-加权马尔可夫链的大坝变形监控模型及预报研究  

Dam Deformation Monitoring Model Based on Gray Neural Network-Weighted Markov Chain and Prediction Research

作　　者：何启[1,2] 戴波[1,2] 

机构地区：[1]河海大学水利水电学院,南京210098 [2]河海大学水文水资源与水利工程科学国家重点实验室,南京210098

出　　处：《中国农村水利水电》2016年第10期146-150,155,共6页China Rural Water and Hydropower

基　　金：国家自然科学基金重点项目(51139001);高等学校博士学科点专项科研基金(20130094110010);病险坝除险加固效应多源多尺度跟踪监控与评析(51179066)

摘　　要：大坝蓄水初期监测资料有限且波动性大,坝体结构处于适应变形的危险阶段。基于这种情况,结合灰色理论、神经网络和加权马尔可夫链理论的优点,构建灰色神经网络-加权马尔可夫链的大坝变形监控模型并对某碾压混凝土坝蓄水初期水平位移进行预测。结果表明,该模型解决了样本数据少、波动性大的问题,拟合效果较好,预测精度较高。The dam monitoring data is limited and the structure adapts to the risk of deformation at the early impoundment stage. In such a situation, by means of the Grey Theory, Neural Networks and Weighted Markov Chain Theory, a Dam Deformation Monito- ring Model based on Gray neural network -- Weighted Markov Chain is built to predict a RCC dam horizontal displacement at the early impoundment stage. Prediction results show that this model solves the problem of fewer data and greater deformation data vola- tility to achieve good fitting and precision.

关 键 词：灰色神经网络 加权马尔可夫链 少数据预测 变形预测 

分 类 号：TV698.1[水利工程—水利水电工程]