关键词: 测井, 含水饱和度, 人工神经网络, 权重, 节点

Abstract: Compared with the traditional comprehensive interpretation and digital processing well logging, BP artificial neural network methods have overwhelming advantages and suitability, it requires no well logging interpretation equation based on testing which the traditional interpretation and processing methods does, it needs the original logging and the practical data for solution, only the original logging data is enough in the process of
the parameter prediction, which provides a new and practical approach to well logging comprehensive interpretation and digital processing for the complicated formations and specials.

Key words: well logging, water saturation, artificial neural network, training, weight, node