摘要
针对传热学反问题的共轭梯度法存在的主要问题,通过在反演过程目标函数中引入正则化项,建立了一种求解传热学反问题的正则化共轭梯度法反演算法。在求解传热学反问题过程中,采用L曲线法确定正则化因子,并采用共轭梯度法对待反演的传热边界条件进行修正。利用数值仿真试验讨论了二维稳态传热系统边界温度分布的正则化共轭梯度反演问题,并与共轭梯度法反演算法进行了比较。结果表明,所建立的正则化共轭梯度法反演算法能够有效抑制测量误差对反演结果的影响。
For the main problems of conjugate gradient method （CGM） used for solving the inverse heat transfer problems, the regularization conjugate gradient method （RCGM） was proposed by introducing the regularization term to the objective function. In the process of solving the inverse heat transfer problems, the L-curve method was used to determine the regularization factor, and the conjugate gradient method was adopted to correct the unknown boundary conditions to be imagined. The numerical simulation was conducted to study the inverse problem of determining the unknown boundary temperature of a two-dimensional steady-state heat transfer system using the RCGM, and the comparison with the CGM was made. The results show that, the RCGM could effectively suppress the impact of measurement error on the estimated results.
出处
《中国电机工程学报》
EI
CSCD
北大核心
2013年第17期78-82,15,共5页
Proceedings of the CSEE
基金
国家自然科学基金项目(51176211)~~
关键词
传热
反问题
温度分布
正则化
共轭梯度
heat transfer
inverse problems
temperature distribution
regularization
conjugate gradient
