高频电磁波测井资料反演的优选调参子空间方法

Optimization Method of Parameter Adjustment Subspace for High-frequency Electromagnetic Well Logging Data Inversion

提出了高频电磁波测井资料反演的一种优选调参子空间方法,给出了调参子空间的优选条件,根据每次迭代反演的计算结果可自动选择调参子空间,将高维反演化为低维反演并使这种低维子空间仍是个好的搜索空间,使迭代反演的收敛速度和精度同时得到提高。对典型地层模型的反演结果表明,优选调参子空间方法是有效的,所给出的调参子空间优选条件是可行的。

This paper proposes an optimization method of parameter adjustment subspace for high-frequency electromagnetic well logging data inversion,providing optimization condition for parameter adjustment subspace.According to each computation result of iterative inversion,parameter adjustment subspace can be automatically selected.Good search space may still exist in low dimension space inversed from high dimension,which lead to simultaneous rising of both convergence rate and convergence accuracy for iterative inversion.Inversion results of two typical stratigraphic models indicate that,optimization method of parameter adjustment subspace,as well as optimization condition for parameter adjustment subspace,brought forth in this paper,both are effect and feasible.