摘要
针对室内试验所得混凝土大坝热学参数的随机性以及传统反分析优化算法易陷入早熟,效率低和计算量巨大等缺点,提出基于并行粒子群算法的反分析方法。本文利用实际所测温度对热学参数进行反分析,根据反分析后的热学参数利用有限元进行温度场正分析,分析比较预测参数与试验参数所得非稳定温度场。结果表明:计算结果和实测结果较为接近,该方法具有较好的稳定性和收敛性,反演结果可满足工程要求。该方法可为混凝土坝热学参数反分析提供理论探讨意义并具有工程实际应用价值。
Aimed at the unreliability of concrete thermal parameters of dam by laboratory testing and the shortcomings of traditional inverse analysis method of easy to fall into premature,low efficiency and great volume of calculation,the paper presented a new inverse analysis method based on parallel PSO.According to the actual measured temperature to predict the thermal parameters,the temperature field should be analyzed by the thermal parameters after inverse analysis,and comparatively analyzed the unstable temperature fields based on the forecast parameters and test parameters.The results showed that this method can be used to obtain thermal parameters of concrete dam and real engineering application value.
出处
《水资源与水工程学报》
2010年第1期97-100,共4页
Journal of Water Resources and Water Engineering
关键词
混凝土坝
热学参数
反分析
并行计算
MPI
concrete dam
thermal parameters
inverse analysis
parallel compute
MPI
