摘要
为了有效控制农药废水纳滤分离工艺运行,基于DK膜预处理吡虫啉废水的试验数据,采用神经网络算法仿真模拟了纳滤系统去除污染物的过程,建立了纳滤分离动态模型,预测了多影响因素作用下的吡虫啉农药废水中污染物去除规律和实时性动态变化,不仅完善了纳滤分离理论系统,而且模型精度满足应用要求,计算的COD、盐分去除率与实测值的相关系数大于0.99,误差在±4%范围内,为农药废水的有效治理提供了必要的技术支持。
For the effective control of the nanofiltration separation of pesticide wastewater,according to the test data of DK membrane pre-treating Imidacloprid pesticide wastewater, this paper analyzes the simulation process of contamination removal using BP neural network and sets up the nanofiltration separation dynamic model. The model precision meets the application requirement for the correlation coefficient between model calculation and test data： the removal ratio of COD and salt is more than 0.99 and the absolute error is less than ±4%. The model predicts the rule of contamination removal in pesticide wastewater under multi-factor condition and timely dynamic movement, improves the theoretical system of nanofiltration separation and provides the necessary support for the effective treatment of pesticide wastewater.
出处
《安全与环境工程》
CAS
2010年第1期22-25,30,共5页
Safety and Environmental Engineering
基金
江苏省环保厅项目(苏环计[2006]002#)
关键词
纳滤
吡虫啉农药废水
BP神经网络
仿真模型
nanofiltration membrane Imidacloprid pesticide wastewater BP neural network
simulation model
