融合算法在测井曲线识别中的应用研究

发布时间：2018-05-02 14:15

本文选题：融合算法 + 测井曲线　；参考：《浙江大学》2015年硕士论文

【摘要】：大庆长垣油田近几年积累了大量的测井曲线,如何高效、快速和经济的从诸多的测井曲线中来分析油层沉积环境,并指导油气勘探和油田开发,智能化测井曲线识别无疑是一种很好的方法。对于识别研究区块的沉积环境、沉积相、沉积微相和水淹层的分布情况具有重要意义。本文通过研究分析测井曲线沉积微相和水淹层的特点及测井曲线响应特征之间的关系。结合提取的特征值建立了数据库,对Rs-LVQ算法进行了改进,构建了Rs-Ga-LVQ融合算法。并对数据库中20口取芯资料的340个沉积微相小层和550个薄差水淹层进行识别,结果显示改进的算法识别率最高。其中261个沉积微相小层与真实结果一致；407个水淹层小层与实际结果一致,达到比较理想的效果。并证明了Rs-Ga-LVQ融合算法应用于测井曲线识别的可行性,且优于Rs-LVQ融合算法。
[Abstract]:Daqing Changyuan Oilfield has accumulated a large number of logging curves in recent years. How to efficiently, quickly and economically analyze reservoir sedimentary environment from many logging curves, and guide oil and gas exploration and oilfield development, Intelligent logging curve recognition is undoubtedly a good method. It is of great significance to identify the sedimentary environment, sedimentary facies, sedimentary microfacies and the distribution of water-flooded layers in the study block. In this paper, the characteristics of sedimentary microfacies and water-flooded layers in log curves and the relationship between the response characteristics of logging curves are analyzed. The database is established based on the extracted eigenvalues, the Rs-LVQ algorithm is improved, and the Rs-Ga-LVQ fusion algorithm is constructed. The 340 sedimentary microfacies layers and 550 thin and differential water-flooded layers of 20 coring data in the database are identified. The results show that the improved algorithm has the highest recognition rate. Of these 261 sedimentary microfacies layers are consistent with the true results; 407 water flooded layers are consistent with the actual results and the results are satisfactory. It is proved that Rs-Ga-LVQ fusion algorithm is more feasible than Rs-LVQ fusion algorithm in logging curve recognition.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P631.81;TE319

【参考文献】