采用BP神经网络模型预测油田废水混凝处理效果  

Predication of the Efficiency of Coagulating Treatment for Oil Field Sewage by BP Neural Network

作　　者：王仁雷[1] 林衍[1] 

机构地区：[1]重庆大学资源及环境科学学院,重庆400044

出　　处：《广州环境科学》2005年第1期6-8,21,共4页Guangzhou Environmental Science

摘　　要：采用均匀设计方法选择神经网络的训练样本,并建立了基于BP人工神经网络的油田废水混凝处理系统的数学模型。利用该模型对实验数据进行预测,结果表明模型的预测值与实测值的平均绝对误差仅为0.74%。可见这种方法预测精度高,具有一定的实际应用价值。A mathematic model based on BP artificial neural network was established for the coagulating treatment of oil fieldsewage,and uniform design was used to choose samples for training the neural network. With this model,the average absolute error be-tween the predicted data and experimental data was only 0.74%,indicating that the model has high predicting accuracy and is worth ofpractical application.

关 键 词：BP神经网络模型 油田废水 混凝处理工艺 水处理技术 混凝剂 非线性回归分析方法 

分 类 号：X741.03[环境科学与工程—环境工程]