摘要
为探究RBF神经网络对青藏工程走廊带冻融砂土导热系数预测效果,采用瞬态平面热源法测试了362个试样导热系数。建立了以含水率和干密度为输入因子的冻融砂土导热系数预测模型。结果表明RBF神经网络可有效预测青藏工程走廊带内冻融砂土导热系数,为冻融土导热系数预测提供了新的思路。
In order to explore the prediction effect of RBF(Radial Basis Function) neural network on the thermal conductivity of frozen and thawed sand in the Qinghai-Tibet Project corridor zone, the thermal conductivity of 362 samples was tested by transient plane heat source method. A model for predicting the thermal conductivity of freeze-thaw sand with water content and dry density as input factors is established. The results indicate that the RBF neural network can effectively predict the thermal conductivity of freeze-thaw sand in the corridor belt of the Qinghai-Tibet Project, which provides a new idea for the prediction of the thermal conductivity of freeze-thaw soil.
作者
刘志云
王伟
崔福庆
张伟
LIU Zhiyun;WANG Wei;CUI Fuqing;ZHANG Wei(School of Geology Engineering and Geomatics,Chang’an University,Xi’an 710054,China;CCCC First Highway Consultants Co.,Ltd.,State Key Laboratory of Road Engineering Safety and Health in Alpine and High Altitude Areas,Xi’an 710075,China)
出处
《低温建筑技术》
2020年第8期95-98,共4页
Low Temperature Architecture Technology
基金
国家自然科学基金(51574037,41502292)
中国交通建设股份有限公司应用基础研究(2018-ZJKJ-PTJS03)。
关键词
青藏高原
多年冻土
RBF神经网络
砂土导热系数
Qinghai-Tibet Plateau
permafrost
RBF neural network
sand thermal conductivity
