正则化方法及其在水电机组动态载荷识别中的应用

项目名称： 正则化方法及其在水电机组动态载荷识别中的应用

项目编号： No.11202116

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： 王林军

作者单位： 三峡大学

项目金额： 24万元

中文摘要： 本项目拟开发一套基于反问题建模和数值反演分析联合的正则化识别技术，以方便、有效地求解复杂工程问题的水电机组动态载荷识别问题。本项目拟从四个方面展开研究，一是高适应度的近似模型技术的研究，重点解决适合于水电机组系统复杂载荷参数识别的反分析模型的构造、非线性方程组的高效迭代求解、反问题的模型处于"超定"状态等难点。二是提出一种稳定求解病态反问题的正则化方法，难点是需要解决正则化算子的构造和正则化参数的选择问题；三是建立适合于水电机组的非线性反问题基本理论与方法，此方法兼有考虑到水电机组系统的高度非线性和不确定性的优点。四是结合反问题建模、数值反演分析与正则化方法，构造高效智能的计算反演算法，并且与自适应管理模型技术有机结合，进一步提高反演的计算效率。

中文关键词： 动态载荷识别；反问题；不适定；正则化方法；水电机组

英文摘要： The project intends to develop a set of regularization recognition technology based on the combination of the inverse modeling and the analysis of numerical inversion in order to conveniently and effectively solve the dynamic load identification of hydropower units in complex engineering problems. The project intends to study four aspects. Firstly, the study of high fitness approximation modeling techniques, which focuses on solving such difficulties as the construction of inverse analysis model being suitable for complex load parameter identification of hydropower unit structure, the efficient iterative solution of nonlinear equations and the model of the inverse problem under the status of "overdetermined"; secondly, we propose the regularization method for a stable solution of the ill-conditioned inverse problem, and its difficulty is to solve the construction of the regularization operator and the selection of regularization parameter; thirdly, we establish the basic theories and methods of nonlinear inverse problems which have the advantage of taking into the highly nonlinearity and uncertainty of hydropower units system ; and fourthly, by the combination of the inverse modeling, numerical inversion analysis and regularization methods, we construct high efficient and intelligent computational inversion

英文关键词： Dynamic load identification；Inverse problems；Ill-posed；Regularization method；Hydroelectric units