摘要
由于在大坝混凝土施工时一些坝段的接缝灌浆区没有预埋温度计,故进行二期冷却时难以判断混凝土灌浆区的温度是否达到接缝灌浆的要求。对此,本文以冷却水管进口温度、出口温度、通水流量、当天坝块的起始温度为输入量,实测混凝土温度为输出量,建立二期通水冷却统计模型,采用BP神经网络对二期冷却时通水工况和坝体混凝土温度作为样本进行训练,然后通过二期冷却时通水工况预测坝体混凝土的温度。实例分析表明,此方法可行。
As there no any thermometers are pre-embedded in the joint grouting zones of some dam sections during concrete construction of dam;it is difficult to judge whether the temperature of the grouting zone meets the requirement of the joint grouting or not when carrying out the second-stage water cooling.For this,a statistical model of the second-stage water cooling is established herein by taking the inlet and outlet temperatures,cooling water flow-rate and the initial temperature of the temperature of intraday dam block as the inputs and taking the measured temperature of concrete as the output,and then the water-flowing condition and the dam concrete temperature during the second-stage water cooling are trained as the specimens with the BP-neural network,thereinafter,the dam concrete temperature is predicted based on the water flowing condition of the second-stage water cooling.The analysis made on an actual case shows that this method is feasible.
出处
《水利水电技术》
CSCD
北大核心
2012年第2期50-53,共4页
Water Resources and Hydropower Engineering
基金
国家自然科学基金项目(51079079)
三峡大学基金(KJ2010B003)
关键词
混凝土坝
二期通水冷却
BP神经网络
温度预测
接缝灌浆
混凝土灌浆区
concrete dam
second-stage water cooling
BP-neural network
temperature prediction
joint grouting
concrete grouting zone
