张玎冉文琼等：应用人工神经网络识别水淹层，测井技术，１９９６（３）２０，２１０～２１４。油层水淹后，水淹层的内部物性明显区别于油层的原始状态。由于水淹状况复杂多变，使用一般测井解释方法识别水淹层具有很大困难。本文研究了应用人工神经网络ＢＰ算法处理测井资料来识别水淹层的基本原理和具体实现，给出了ＢＰ算法的改进方法及在两个油田的实际应用效果。 Zhang 玎 Ran Wenqiong et al: Application of Artificial Neural Network to Identify Flood Flooded Layer, Logging Technology, 1996 (3) 20,210-214. After the reservoir is flooded, the internal physical properties of the flooded layer are obviously different from the original state of the oil layer. Due to the complexity and variety of flooding conditions, it is very difficult to identify the flooded layer using the common logging interpretation method. In this paper, the basic principle and concrete realization of using BP neural network to process well logging data to identify water flooded layer are studied. The improved BP algorithm and its application in two oil fields are given.