论文部分内容阅读
将神经网络应用于油田注水开采系统的动态指标预测。应用某油田１９８１～１９９３年１３年开发系统典型因素历史数据，对建立起的ＢＰ网络模型进行训练。训练后的网络既能准确地拟合历史数据，又能较精确地预测油田动态指标。
Application of Neural Network in Prediction of Dynamic Indexes of Oilfield Water Injection System. Applying the historical data of typical factors of a 13-year development system of a certain oilfield from 1981 to 1993, we train the established BP network model. The trained network can not only accurately match the historical data but also predict the field dynamic index more accurately.