摘要
从油气储层类型识别的现状出发,提出采用弹性BP神经网络进行油气储层类型识别的原理、方法。与试油试水资料比对的结果表明,弹性BP神经网络可在不增加额外测井数据的情况下,对储层类型进行较准确的判别,相对于常规BP网络,可在减少网络训练时间的同时,提高网络预测的正确性。
This paper puts forward the theories and methods for recognizing the types of oil and gas reservoirs by resilient BP neural network in terms of the present situation of type-recognizing of oil and gas reservoir. The comparison of water-testing and oil-testing data shows that the resilient BP neural network can make more accurate discrimination of the types of reservoirs without any additional data of well logging. Compared with the regular BP network,it can improve the correctness of network prediction while shortening the time of network experiment.
出处
《科技情报开发与经济》
2010年第6期142-144,共3页
Sci-Tech Information Development & Economy
关键词
储层类型
BP神经网络
弹性算法
types of reservoir
BP Neural network
resilient algorithm
