摘要
采用不同方法对基于热传导反问题的固体热导率预测进行了研究。分别采用Bayesian统计方法、Levenberg-Marquardt和遗传算法对二维各向异性材料的热物性进行了预测,并进行了分析比较。研究结果表明,Bayesian方法中热传导反问题的解是其后验概率密度的数学期望,而后验概率密度函数(PPDF)通过测定的温度进行计算获得,用Markov chain Monte Carlo算法计算后验状态空间以得到未知热导率的统计估计,采用Me-tropolis-Hasting算法进行数据采样构造Markov chain,并截取收敛后的样本进行分析。遗传算法是一种相对较新的用于最优化估计的方法,也可以用于求解反问题。
Several inverse analysis methods were used to estimate linearly dependent thermal conductivity components of an orthotropic solid. The solutions of inverse heat conduction problems by using Bayesian approach, genetic algorithms and Levenberg-Marquardt algorithm were compared. The inverse solution of estimating thermal conductivity components were obtained by computing the expectations of the posterior probability density functions of the thermal conductivity components. Genetic algorithm is a relatively new technique for optimization and estimation. This technique can also be applied to solving inverse problems. Levenberg-Marquardt algorithm can also be used to solve the nonlinear inverse problem of the parameter estimation.
出处
《化工学报》
EI
CAS
CSCD
北大核心
2007年第6期1378-1384,共7页
CIESC Journal
基金
国家自然科学基金项目(50576106)。~~
