基于模糊C均值聚类的沉积相定量识别——以川西某气田蓬莱镇组为例

利用传统定性的方法识别沉积微相，很大程度上依赖于研究人员的经验、识别效率低；通过岩心观察、粒度分析，沉积构造特征和测井相分析，结合区域地质背景，首先确定出川西某气田蓬莱镇组的沉积微相类型为浅水三角洲沉积体系，主要发育三角洲前缘亚相的水下分流河道、河口砂坝、席状砂、水下分流间湾及席间泥等微相；然后根据目的层段的分层数据表提取测井曲线（自然电位、自然伽马、声波时差及地层电阻率等），并计算反映沉积微相特征的参数（平均值、地层厚度、相对重心及变差方差根等参数）；最后对这些参数进行归一化处理及模糊C均值聚类识别，结合沉积微相的平面分布规律，绘制出该层的沉积微相；基于此编制了相应的识别程序，对研究区的11个主力小层进行了沉积微相识别；与常规方法相比，该法可以批量识别沉积相，弥补了传统定性识别沉积微相的缺陷，识别的准确率较高，具有一定的推广价值。

Quantitative Identification of Sedimentary Facies Based on Fuzzy C-means Clustering Algorithm: A Case Study from Penglaizhen Formation in the X Gasfield in Western Sichuan

The traditional method of qualitative identification of sedimentary microfacies always depends largely on the researchers’ experience with relatively low identification efficiency.Fuzzy C-means clustering algorithm is applied to the quantitative identification of sedimentary microfacies within the Penglaiba Formation in the X Gasfield,Western Sichuan,it includes the following steps.First,on the basis of core observation,particle size analysis,and log facies analysis,together with regional geological background data,it was concluded that the sedimentary microfacies type of the Penglaiba Formation in the study area is dominated by shallow-water delta depositional system,including five microfacies belongs to deltaic front facies such as submerged distributary channel,mouth bar,sand sheet,submerged distributary interchannel,and inter-sand sheet shale.Then some characteristic parameters(such as VA,h,W and GS) representing sedimentary microfacies were calculated by well logs(SP,GR,AC and RT),which are extracted according to the hierarchical data table of main target series of layers.After normalizing these parameters,the quantitative identification of sedimentary microfacies by using fuzzy C-means clustering algorithm were carried out.Finally,the sedimentary microfacies is delineated for this layer with the help of its plane distribution law.For convenience operation,a recognition program is developed,helping the identification of sedimentary microfacies for some other eleven production sub-layer in the study area.Compared with conventional method,this method is mentioned here can be used for batch recognition of sedimentary facies,making up for the defects of traditional qualitative identification of sedimentary microfacies with a relatively higher Recognition accuracy.So this method has some promotional value.