煤层重复压裂选井选层的数学模型研究

发布时间：2018-05-24 14:56

本文选题：煤层气 + 重复压裂　；参考：《西南石油大学》2017年硕士论文

【摘要】：无论低产的煤层气井还是减产的煤层气井,都必须进行重复压裂;尽管目前国内煤层气开发还主要处于一次压裂阶段,但煤层重复压裂是未来的趋势和主流。而煤层重复压裂选井选层是煤层重复压裂的基础,也是事关煤层重复压裂效果的关键和难题,研究必要,意义重大。本文研究煤层重复压裂选井选层的数学模型,以影响煤层重复压裂效果的因素分析为基础,分别针对训练样本少和多的情形,建立相应的数学模型。经过硕士期间的研究工作,本文获得以下成果:(1)通过对原始数据进行预处理,综合应用灰色关联度分析、相关性分析和主成分分析,揭示了影响煤层重复压裂效果的主控因素,并奠定了建立煤层重复压裂选井选层数学模型的基础。(2)针对训练样本少的情形,以影响煤层重复压裂效果的主控因素为基础,建立了煤层重复压裂选井选层的模糊综合评价模型。(3)针对训练样本多的情形,以影响煤层重复压裂效果的主控因素为基础,建立了煤层重复压裂选井选层的概率神经网络模型,突出考虑了参数取值的随机性。
[Abstract]:No matter the coalbed methane wells with low yield or those with reduced production, they must be fracked repeatedly. Although the development of coalbed methane in China is still mainly in the stage of single fracturing at present, the repeated fracturing of coal seam is the trend and mainstream in the future. The selection of wells and layers by repeated fracturing of coal seam is the basis of repeated fracturing in coal seam, and it is also the key and difficult problem related to the effect of repeated fracturing in coal seam. It is necessary and significant to study. In this paper, the mathematical model of selecting wells and layers by repeated fracturing in coal seam is studied. Based on the analysis of factors affecting the effect of repeated fracturing in coal seams, the corresponding mathematical models are established for the case of less and more training samples. Through the research work during the master's period, this paper obtains the following achievements: 1) by preprocessing the original data, synthetically applying grey correlation analysis, correlation analysis and principal component analysis, this paper reveals the main controlling factors that affect the effect of coal seam refracturing. The foundation of establishing mathematical model of selecting wells and selecting layers by repeated fracturing in coal seam is established. Aiming at the situation of few training samples, the main controlling factors that affect the effect of coal seam refracturing are taken as the foundation. A fuzzy comprehensive evaluation model for selecting wells and layers by repeated fracturing in coal seam is established. Aiming at the situation of a large number of training samples, a probabilistic neural network model for selecting wells and layers by repeated fracturing in coal seam is established on the basis of the main controlling factors that affect the effect of repeated fracturing in coal seam. The randomness of parameter values is considered prominently.
【学位授予单位】：西南石油大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TE377;O159;TP183

【相似文献】