论文部分内容阅读
考虑了一种五层结构的正规化模糊神经网络模型,针对网络结构的优化问题给出了该网络模型的规则层节点的选取方法和相应的反传播学习规则。对于具有较少数输入节点的情况,该网络有较快的训练速度。油藏测井解释中水淹层的识别是石油开发中特别是开发中后期比较突出的一个问题,复杂的地质条件在测井曲线的表现中具有许多模糊性,在各种模糊条件的组合下油藏水淹表现为强水淹、中水淹、弱水淹和无水淹等情形。将正规模糊神经网络用于油藏测井解释中水淹层的识别以提取测井曲线与水淹级别之间的映射关系,从而实现模糊性油藏测井解释中水淹层的识别。实验表明此方法对解决水淹层识别问题具有良好的适应性和实用性。
A five-layer normalized fuzzy neural network model is considered. For the optimization problem of the network structure, the selection method of the node in the rule layer of the network model and the corresponding anti-propagation learning rule are given. For networks with fewer input nodes, the network has a faster training speed. The identification of water-flooded layer in reservoir logging interpretation is a prominent issue in the oil development, especially in the middle and late stages of development. Complicated geological conditions have many ambiguities in the performance of well logging curves. Under the combination of various fuzzy conditions The reservoir flooding is characterized by strong flooding, flooding, weak flooding and no flooding. The formal fuzzy neural network is used to identify the water flooded layer in reservoir log interpretation to extract the mapping relationship between well logging curve and flooding level so as to realize the recognition of water flooded layer in logging interpretation of ambiguous oil reservoir. Experiments show that this method has good adaptability and practicability to solve the problem of water flooded layer identification.