油井流动图像重建算法研究  

Study on Flowing Image Reconstruction Algorithms for Oil Wells

作　　者：别静[1] 吴锡令[1] 缪志伟[2] 

机构地区：[1]中国石油大学(北京)地球物理与信息工程学院,北京100029 [2]中国石化勘探南方分公司,四川成都610041

出　　处：《石油天然气学报》2011年第7期92-94,3+2,共3页Journal of Oil and Gas Technology

基　　金：国家自然科学基金项目(40114308)

摘　　要：非迭代和迭代电磁流动成像(FEMT)图像重建算法分别具有成像质量差和成像速度慢的缺点。针对这一问题,提出了一种基于Tikhonov正则化的修正共轭梯度图像重建算法。该算法对传统共轭梯度算法的搜索方向进行了修正,同时采用Tikhonov正则化算法作为其起始步。成像结果表明,该算法在图像重建速度和质量上都优于传统的共轭梯度算法。Non-iterative and iterative image reconstruction algorithms for flowing electromagnetic tomography(FEMT) had their drawbacks of low accuracy and slow imaging.To solve the problems,a modified conjugate gradient algorithm based on Tikhonov regularization was proposed.The new method was used to modify the search direction of the traditional conjugate gradient algorithm,and Tikhonov regularization was used as its first step.Imaging results show that the algorithm can provide high performance in image reconstruction speed and quality compared with the traditional conjugate gradient algorithm.

关 键 词：流动成像 修正共轭梯度算法 TIKHONOV正则化 图像重建 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]