摘要
Fisher判别根据投影原理将多个变量的特征差异表现在几个主要的判别函数中,具有算法稳健、信息量大、准确率高和应用方便的特点,适用于多总体多变量的火山岩岩性识别研究。在研究了准噶尔盆地红山嘴油田石炭系火山岩岩石学特征和3亚类9小类岩性测井响应特征的基础上,用Fisher判别分别建立了相应的火山岩岩性识别模型,判别率较高。用该成果对本区46口井的常规测井资料进行了岩性识别,为火山岩油藏难采储量评价奠定了基础。
Lithologic identification is fundamental technology in volcanic reservoir research. Fisher discrimination can be used to represent multiple variances differences into several functions, with algorithm stability, massive information, high accuracy and convenient application, so generally applied to volcanic lithologie identification researches. Base on petrologic characters of volcanie rocks of Carboniferous in Hongshanzui oilfield of Junggar basin and investigation on log responses of three sub-types and nine kinds of rocks, this paper establishes the identifieation model of those rocks with high recognition rate using Fisher discrimination. Application of this result to 46 wells with regular logging data for lithologic identification is made, laying a foundation for evaluation of difficult-recovery reserves of such a volcanic reservoir in this area.
出处
《新疆石油地质》
CAS
CSCD
北大核心
2008年第6期761-764,共4页
Xinjiang Petroleum Geology
关键词
红山嘴油田
火山岩
岩性识别
石炭系
Hongshanzui oilfield
volcanic rock
lithologic identification
Carboniferous