摘要
为实现对造纸废水的有效控制,提升造纸企业的环保生产能力,基于纸张生产废水的主要成分建立了一套以温度、PH值、进水COD和进水流量为基础指标的反向传播神经网络模型,通过该网络输出造纸工艺的出水COD和产气量。为实现针对造纸废水的多目标优化,在反向传播神经网络模型的基础上引入了NSGA-Ⅱ算法,最终形成了一套多目标废水处理优化模型。为验证多目标废水处理优化模型的有效性,基于造纸企业所提供的废水成分数据对出水COD和产气量两项指标进行预测,发现该模型所输出的预测结果接近真实值,适用于造纸废水厌氧处理工作,具有一定的应用价值。
In order to achieve effective control of papermaking wastewater,improve the environmental protection production capacity of papermaking enterprises.Based on the main components of paper production wastewater,a set of backpropagation neural network models based on temperature,pH,influent COD and influent flow were established,and the effluent COD and gas production of the papermaking process were output through the network.In order to realize the multi-objective optimization of papermaking wastewater,the NSGA-II algorithm was introduced on the basis of the backpropagation neural network model,and finally a set of multi-objective wastewater treatment optimization models were formed.In order to verify the effectiveness of the multi-objective wastewater treatment optimization model,the two indicators of effluent COD and gas production were predicted based on the wastewater composition data provided by papermaking enterprises,and it was found that the prediction results output by the model were close to the real value,which was suitable for the anaerobic treatment of papermaking wastewater and had certain application value.
作者
宋文雅
王可涵
SONG Wenya;WANG Kehan(Xi’an Aeronautical Vocational and Technical College,Xi’an 710021,China)
出处
《造纸科学与技术》
2024年第9期43-45,共3页
Paper Science And Technology
基金
陕西省教育厅2022年度一般专项科研基金项目(22JK0104 2022HZ1188)。
关键词
反向传播神经网络
多目标优化模型
NSGA-Ⅱ算法
仿真试验
backpropagation neural network
multi-objective optimization mode
NSGA-II algorithm
simulation experiment
