论文部分内容阅读
近年来,在物探资料推断解释过程中不断引入一些最优化反演算法.无约束最优化算法往往因为多解性,效果一般都不太好.所以最近在地球物理反演理论方面,已逐渐注意研究在约束条件下的最优解问题.常用的数学工具是线性或非线性规划(带约束)方法,本文着重讨论其中的惩罚函数法,并把它引入磁异常模型和实例的反演计算.计算结果表明,它确能减少反演解的多解性,比一般无约束方法能取得更有意义的结果.
In recent years, some optimization inversion algorithms are continuously introduced in the process of deducing and explaining geophysical data.Unconstrained optimization algorithms are usually not good due to their multiple solutions, so in the field of geophysical inversion recently, attention has been paid We study the optimal solution under constrained conditions.The commonly used mathematical tools are linear or nonlinear programming (with constraints) method, this article focuses on the penalty function method and introduces it into the calculation of magnetic anomaly model and instance inversion. The calculation results show that it can indeed reduce the multi-solution of the inversion solution, and can achieve more meaningful results than the general unconstrained method.