交会图和BP神经网络技术在碎屑岩识别中的应用

饶阳凹陷新近系馆陶组岩性以碎屑岩为主,岩性复杂多样,单纯利用测井曲线难以对岩性进行较好地识别,对后续测井解释的结果造成了不利的影响。针对该问题,以测井资料为基础,提出了一种首先利用交会图技术将各类岩性进行归纳总结,然后应用BP神经网络技术对归纳后的岩性进行快速识别的方法。从此种方法在留西地区的应用效果来看,该方法对样本数据库中各类岩性的识别精度达到了90%以上,其中泥岩和粉砂岩的识别精度更是达到了100%,大大提高了单纯利用测井曲线对岩性进行分类识别的精度,在油气勘探开发过程中能够发挥较为重要的作用。

Application of Crossplot and BP Neural Network Technique in the Identification of Clastic Rock

The lithology of Raoyang depression Neogene Guantao group mainly is clastic rocks,the lithol-ogic characteristic is complicated and varied of which better identification is hard to be achieved only using the logging curve,it has unfavorable influence on subsequent well logging interpretation.Targeting at this problem and based on the logging data to propose a method which first using the crossplot technique to generalize and summarize all kinds of lithologies and then using the BP neural network technique to quick i-dentify the generalized lithologies.According to the application effect of this method in Liuxi area,the iden-tification precision of all kinds of lithologies in the sample data base has reached over 90% and the identifi-cation precision of mudstone and siltstone have even reached 100%,it greatly improve the classification and identification precision of all kinds of lithologies only using the logging curve,it has more importantly in-fluence in the oil and gas exploration and development process.