项目名称: 介质热传导反问题的正则化方法及数值解
项目编号: No.11201066
项目类型: 青年科学基金项目
立项/批准年度: 2013
项目学科: 数理科学和化学
项目作者: 闫亮
作者单位: 东南大学
项目金额: 22万元
中文摘要: 由于工业生产的推动,介质热传导问题受到了广泛关注。实际热传导问题所能提供的数据往往是不完全的,带有很大的随机误差,同时相应问题还具有非线性以及严重不适定性,因此经典的数值方法处理这些问题将变得十分复杂,稳定性难以保证。对这类非线性不适定模型进行数学理论分析,结合有效的正则化方法提出和发展新的高效稳定算法是目前热传导反问题及其应用中迫切需要解决的问题。本项目将围绕钢铁生产中的介质热传导反问题开展两个方面的研究。一是研究处理介质热传导反问题的无网格方法,发展该方法在处理介质热传导反问题尤其是关于二相界面重建以及高维热传导参数识别问题中的有效应用,拟得到一些突破性成果。二是研究处理介质热传导问题参数识别的贝叶斯推断方法,旨在贝叶斯框架下建立相应的理论结果以及数值算法的快速实现,进一步提高该方法的有效性和实用性。通过这些问题的研究,给出解决介质热传导反问题的有效数值计算方法及理论分析。
中文关键词: 反问题;无网格方法;贝叶斯推断;分数阶微分方程;随机替代模型
英文摘要: Inverse heat conduction problems(IHCP) deal with the estimation of unknown quantities appearing in the mathematical formulation of physical processes in thermal sciences, by using measurements of temperature, heat flux, radiation intesities, etc. Nowadays, inverse analyses are commonly encountered in single- and multimode heat transfer problems which arise from diverse industrial applications and their requirement, dealing with multiscale phenomena. Applications range from the estimation of constant heat transfer parameters to the mapping of spatially varying and time-varying functions, such as heat sources, fluxes, and thermophysical properties. Those problems are mathematically and numerically very challenging due to their inherent ill-posedness and nonlinearity. Therefore, special numerical techniques are required for the solution of inverse problems. This project attempts to design some effcient algorithms based on both Bayesian statistical approached and deterministic regularization methods for the inverse heat conduction problems. The specific inverse problem under consideration arises in steel producing process. It consists of two parts: Part I discusses the application of the meshless method for the IHCP, especially in the two-phase Stefan problems and high dimension parameter estimation problems, whi
英文关键词: inverse problems;meshless method;Bayesian inference;fractional PDEs;Stochastic surrogate model