高拱坝施工期垂直向温度分布智能动态预测  

Intelligent Dynamic Prediction about the Vertical Temperature Distribution of High Arch Dam during Construction Period

作　　者：唐腾飞[1] 黄耀英[2] 周宜红[2] 郝鹏[1] 王飞[1] 

机构地区：[1]中国电建集团贵阳勘测设计研究院有限公司,贵州贵阳550081 [2]三峡大学水利与环境学院,湖北宜昌443002

出　　处：《水力发电》2015年第12期54-57,共4页Water Power

基　　金：国家自然科学基金(51079079;51209124)

摘　　要：针对当前特高拱坝施工期混凝土温控采取分区分时段通水冷却难以实时掌握并准确预测垂直向温度分布的问题,利用分布式光纤实时在线监测采集数据量大的优势,发掘其线空间内的数据关系,基于广义回归神经网络建立隐式垂直向温度分布模型,并对某处于建设期的高拱坝垂直向温度分布进行动态预测分析。结果表明,建立的垂直向温度预测模型的预测温度与实测温度之间存在良好的相关性,能很好地反映各温控因素与监测数据的非线性关系,能够准确、合理地预测垂直向温度。The concrete temperature control in construction period of high arch dam always adopts water cooling in different time and space, but it is difficult to grasp and predict the vertical temperature distribution. By using abundant online monitoring data obtained from distributed optical fiber, the relationship of data in line space is explored based on generalized regression neural network to set up an implicit expression model about the vertical temperature distribution. The model is used to predict the vertical temperature distribution of a high arch dam in construction period. The results show that the predicted temperature of model has a good relation with actual online monitor temperature and it can reflect the non-linear relationship between temperature control factors and monitoring data. The model can accurately and reasonably predict vertical temperature distribution.

关 键 词：高拱坝 施工期 广义回归神经网络 垂直向温度分布 智能动态预测 

分 类 号：TV698.1[水利工程—水利水电工程] TV642.4