南水北调与水利科技Issue(2)：354-357,4.DOI:10.13476/j.cnki.nsbdqk.2015.02.035
基于BP神经网络的船闸闸首热学参数反分析
Back analysis on thermal parameters of ship lock head based on BP neural network
苏超 1王璐 1董义佳 1陶文兴2
作者信息
- 1. 河海大学 水利水电学院，南京 210098
- 2. 越南水利大学，越南 河内 100803
- 折叠
摘要
Abstract
Due to engineering construction limits, the thermal parameters of concrete for some lock head projects cannot be deter-mined by the adiabatic temperature rising test. In this paper, an inversion analysis method was proposed to determine the lock head thermal parameters based on BP neural network. First, the combinations of concrete thermal parameters were constructed based on the uniforMdesign theory and they were used to generate a series of training samples using the thermal FEManalysis. Then,themeasuredtemperaturedatawereputintotheneuralnetworktoperformtheinversionanalysisonthethermalparame-ters.Final y,the thermal parameters obtained froMthe inversion analysis were used to analyze and predict the temperature data atcertainpointsandcomparedthemwiththeactualtemperaturedata.TheresultsshowedtheconvergencerateoftheBPneural network is improved by the thermal parameters developed by the uniforMdesign, and the obtained parameters meet the engi-neering requirements.关键词
船闸闸首/热学参数/反分析/BP 神经网络/三维有限元Key words
lock head/thermal parameters/back analysis/BP neural network/three-dimensional finite element method分类
交通工程引用本文复制引用
苏超,王璐,董义佳,陶文兴..基于BP神经网络的船闸闸首热学参数反分析[J].南水北调与水利科技,2015,(2):354-357,4.
