非线性反演算法探究  

The Exploration of Nolinear Inverting Algorithm

作　　者：王岩[1] 

机构地区：[1]吉林建筑工程学院网络信息中心,长春130118

出　　处：《吉林建筑工程学院学报》2011年第3期75-78,82,共5页Journal of Jilin Architectural and Civil Engineering

摘　　要：自20世纪90年代以来,应该说非线性反演方法已经比较成熟.例如,线性化迭代反演、模拟退火法、遗传算法及联合反演方法等.本文就无约束非线性反演的几种方法、约束反演、岭回归或Marquardt-Levenberg法的实现加以说明,并详细阐述了SVD在非线性反演中的应用,对岭回归时阻尼因子的确定以及应用先验信息的非线性反演方法进行了深入的讨论.It is sure that the nonlinear inversion method,such as linear iterative inversion method,simulated annealing method,genetic algorithm and the joint inversion method,etc.,has been more mature since the 1990s.This dissertation elucidates the achievement of several methods of unconstrained-nonlinear inversion,constraint inversion method,ridge regression method or Marquardt-Levenberg method,plus expounds the application of SVD in nonlinear inversion.The confirmation of the damping factor in ridge regression and the nonlinear inversion method using transzendental information are discussed.

关 键 词：非线性反演问题 岭回归法 非线性阻尼最小二乘法 阻尼因子 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]