神经网络自动识别沉积微相在胡状集油田的应用

Nerve Network Automatically identification of Sedimentary Microfacies in Huzhuangji Oilfield

秦亚玲¹,计　平²,郑宇霞³,阎育英³

Qin Yaling, Ji Ping, Zheng Yuxia, et al.

1.中原油田分公司地质调查处,2.中原油田分公司审计一所,3.中原油田分公司勘探开发科学研究院

Geologic Survey and Data Processing Company , Zhongyuan Oilfield Company ,SINOPEC, Henan 457001 , P. R. China

为提高胡状集油田胡十二断块沙三段各沉积时间单元沉积微相划分的准确性,在系统取心井单井相研究的基础上,优选并提取了能反映各种沉积微相特征的定量参数,对不同类型的微相进行定量标定。将标定结果输入人工神经网络,应用神经网络的智能功能,并通过自动识别、调整权值,实现对未知沉积时间单元微相的自动识别。用神经网络方法对胡状集油田150 多口井95 个沉积时间单元进行沉积微相划分,取得了较为理想的结果,避免了仅用测井曲线划分沉积微相的不确定性。

To improving the accuracy of division of each sedimentary microfacies in Hu 12 fault-block in Huzhuangji Oilfield , the single well factes of cored wells were researched , and the optimum quantitative parameters , which can reflect the characteristics of various sedimentary microfactes , were calibrated quantitatively. Inputting the result into the nerve network system , the automatic dentification of unknown sedimentary microfacies were done by the automatic identification and adjustment of weighted value of nerve network’s intelligent function. The nerve network technique were used to divide sedimentary microfacies among more than 150 wells , 95 sedimentary units in Huzhuangji Oilfield , the result is expected