利用人工神经网络预测砂岩储层渗透率  

Predicting reservoir permeability of sandstone by means of artificial neural network

作　　者：张雁[1] 柳成志[1] 秦秋寒[1] 张世广[1] 翟永库[2] 

机构地区：[1]大庆石油学院地球科学学院,黑龙江大庆163318 [2]大庆油田有限责任公司测试技术服务公司,黑龙江大庆163414

出　　处：《大庆石油学院学报》2005年第4期10-11,32,共3页Journal of Daqing Petroleum Institute

摘　　要：根据渗流原理,利用人工神经网络算法建立了预测砂岩储层渗透率的模型,应用该模型分析了萨尔图油田中区储层渗透率.结果表明:预测渗透率与实测渗透率吻合较好,预测渗透率的变化趋势与取心井段岩性的变化规律相符;应用人工神经网络预测渗透率的方法,既可保证数据总体分布的正确性,又可提高渗透率预测的精度.According to the filtration principle, the model of predicting reservior permeability of sandstone was set up using artificial neural network algorithm, the reservior permeability was analyzed with the model in center district of Sartu oil field. The result shows: The permeability predicted and the permeability surveyed proved to be almost identical, the predicted variation tendency of permeability is consistent with lithologic variation tendency in the coring. Accordingly, the method of permeability predicting by means of artificial neural network can already guarantee the exactness of overall data and improve the precision of permeability predicting.

关 键 词：人工神经网络 渗透率 砂岩储层 

分 类 号：TE122.2[石油与天然气工程—油气勘探]