人工神经网络在低阻油藏产状评价上的应用

Appl ication of Artifical Nerve Networks to Evaluation of Reservoir Occurrence of Low Resistivity

 杨庆军 邓春星 卫　东

中国地质大学石油系，中原石油勘探局国际合作部 

Department of Petroleum, China University of Geosciences , Hubei 430074 , P. R. China

低阻油藏是一种非常规油藏,电性与物性之间存在着一种模湖的非线性关系。如何评价它的产状是研究和生产的难点。本文从分析油井的产能出发,导出了影响油藏产状的因素。并利用人工神经网络具有自适应、自学习的特点,将神经网络与常规测井、试油、试井等动态资料相结合进行油藏产状评价,取得了较好的效果。
 

Having analysing the production capacity of oil wells , the factors which affect the reservoir occurrence are concluded. Using the characters of artificial nerve networks ,adapt it self ,learn it self and antijamming ,it is combined with dynamic data ,such as ,conventional logging ,oil production test and well testing ,to evaluate reservoir ,good effect s are received.