一种新的核磁共振弛豫谱反演算法

A new Inversion Algorithm for NMR Relaxation Distribution

提出了一种新的基于整体迭代修正的核磁共振弛豫信号的多指数反演算法 ,对算法的具体实现过程进行了详细的论述 ,并通过算法实例处理与基于奇异值分解法的MAPⅡ算法的反演结果进行了对比 ,对比总结了不同信噪比下、不同布点方式下核磁共振弛豫谱的反演结果 ,表明在 16～ 64点对数均匀布点时SIRT反演结果能保证岩石、流体弛豫谱的真实性 ,该算法对岩心核磁共振实验分析以及核磁共振测井资料解释分析具有重要的参考价值。

A new multi-exponential inversion algorithm for NMR T 1 and T 2 relaxation distributions of rocks and free fluids is presented based on SIRT, Solid Iteration Redress Technique. The concrete implementation process for SIRT is discussed in detail, and some SIRT inversion results are contrasted with MAPⅡ results based on SVD algorithm. The contrasts of inversion NMR spectrums that have different preassigned relaxation times for echoes with different SNR verify that the SIRT algorithm with 16～64 log spaced time constants can give true relaxation distributions for rock core and free fluids, and the new inversion algorithm is a valuable reference for rock core NMR experimental analyses and NMR log interpretations.