摘要
为最大限度提取自旋回波中的有效信息,提出一种新的T2谱反演方法。该方法首先依据噪声的方差自适应选取奇异值,得到零空间矩阵将解向量映射到低维空间,在低维空间建立最小二乘目标函数,将其转化为解非负二次规划问题,采用有效集解法得到最优解。数值仿真实验结果表明,自适应截断奇异值法在准确性、稳定性和抗噪性方面优于固定截断奇异值法,有效集非负约束方法在准确性方面优于传统非负约束方法。该方法保证了T2谱的准确性和稳定性,并且运算速度较快,整体上优于奇异值分解法和最优化法。
This paper proposes a new T2 spectrum inversion method designed for the maximum extraction of the effective information from the spin echo to the maximum.This method operates by selecting singular values adaptively according to the variance of the noise;obtaining a zero-space matrix and mapping the solution vector to a low-dimensional space where a least-squares objective function is established to transform it into a solution to the non-negative quadratic programming problem;and obtaining the optimal solution using effective set solution method.The numerical simulation experiments show that the adaptive truncation singular value method outperforms the fixed truncation singular value method in terms of accuracy,stability,and noise resistance,and the effective set non-negative constraint method is superior to the traditional non-negative constraint method in accuracy.The proposed method capable of a greater accuracy and stability of the T2 spectrum and a faster calculation speed demonstrates an overall advantage over the singular value decomposition method and optimization one.
作者
李鹏举
王梓丞
任莉
Li Pengju;Wang Zicheng;Ren Li(School of Earth Science, Northeast Petroleum University, Daqing 163318, China;Accumulation & Development of Unconventional Oil & Gas, State Key Laboratory Cultivation Base Jointly-Constructed by Heilongjiang Province & Ministry of Science & Technology, Northeast Petroleum University, Daqing 163318, China;Interpretation & Evaluation Center, Daqing branch of CNPC logging Co.Ltd., Daqing 163000, China)
出处
《黑龙江科技大学学报》
CAS
2020年第3期264-269,共6页
Journal of Heilongjiang University of Science And Technology
关键词
岩石
反演方法
自适应截断奇异值
有效集非负约束
奇异值分解
rock
inversion method
adaptive truncated singular value
active set non-negativity constraint
singular value decomposition
