BP神经网络预测废水处理过程的研究  

Studies on predicting the effluent treatment process with BP neural network

作　　者：张燕聪[1] 万金泉[1] 马邕文[1] 

机构地区：[1]华南理工大学造纸与环境工程学院,广东广州510640

出　　处：《中华纸业》2006年第2期60-62,共3页China Pulp & Paper Industry

基　　金：广东省科技厅重大专项攻关项目(编号2003A30406)

摘　　要：通过在实验室条件下进行造纸废水处理试验取得的数据对BP神经网络进行训练,建立了造纸废水处理过程的网络模型。该网络模型仿真实际废水处理过程的结果表明,BP神经网络具有很强的学习能力。利用BP神经网络模型实现了对造纸废水处理过程的预测,平均相对误差为19%,表明网络泛化能力不是很好。经过动态训练的BP神经网络模型能够比较准确的预测造纸废水处理过程,平均相对误差为1.9%,大大优于未经动态训练的网络模型。The BP neural network trained with the data from the papermaking effluent treatment experiment under laboratory conditions and a network model was built for the papermaking effluent treatment process. The results of emulating practical wastewater treatment process of the network model show that the BP neural network has a strong learning capability. The effluent treatment process was predicted with this BP neural network model with the average relative error of 19%, which indicates that the generalization power of the network is not so desirable. While the BP neural network model after a dynamic training, it had a more precise prediction on the papermaking effluent treatment process with the average relative error of 1.9%, which indicates that the dynamic training BP neural network is greatly superior to the one without dynamic training.

关 键 词：BP神经网络 预测模型 造纸废水处理 

分 类 号：X793[环境科学与工程—环境工程] TS756[轻工技术与工程—制浆造纸工程]