摘要
由于核磁共振测井数据的信噪比很低 ,能否很好地降低噪声的影响成为核磁共振测井解谱方法有效与否的关键因素。本文介绍了在解谱中将特征矩阵进行奇异值分解 ,然后截去小的非零奇异值的方法 ,以增加特征矩阵的稳定性 ,从而保证数据在低信噪比时仍然可以得到稳定的弛豫频谱。本方法在实际应用中取得了很好的效果。
Due to the low singal-to-noise ratio of log data from NMR tool,reducing the influece of noise effectively is a key factor in transforming NMR logging data into T 2 relaxation spectra.The paper introduced the SVD of solving eigenmatrix in solution of T 2 relaxation spectra and the feasibility and the method that cuts-off the non-zero singular value in SVD in order to increase the stability of eigenmatrix.All these can ensure that the steady relaxation spectra can be obtained from low singnal-to-noise NMR log data.This method has a good result in practice.
出处
《石油地球物理勘探》
EI
CSCD
北大核心
2003年第1期91-94,共4页
Oil Geophysical Prospecting
关键词
应用
奇异值分解
算法
核磁共振测井
解谱方法
油气勘探
nuclear magnetic resonance log,singular value decomposition,solution of T 2 relaxation spectra
