机构地区: 华中理工大学电子与信息工程系
出 处: 《地球物理学报》 1994年第A02期562-568,共7页
摘 要: 本文提出了一种利用多级BP神经网络进行石油测井信号分类的新方法.介绍了用多级BP网络处理测井信号的分类器算法和网络结构,并给出了针对理论模拟信号的分类结果及针对实际模型井信号的分类结果,其正确率可达90%以上. A new method to classify the oil logging signal using a multistage BP neural network is proposed in this pape The network structure is introduced. The improved back-propagation algorithm employed to train the BP network is presented The experimental results of classifying the theoretical simulative signal and the measured signal from real simulative wells are also introduced. The correct classification rate achived by this neural network is over of 90%.
关 键 词: 网络 信号识别 神经网络 测井 采油井 分类器
领 域: [石油与天然气工程]