摘要
借助于极限学习机的自学习功能,确定适合AAO工艺污水处理厂的最优内外回流比,实现节能降耗和提高出水水质的目标。首先收集污水处理厂一年的监测数据,然后利用ELM建立网络并进行训练与测试,训练结果是隐含层节点数设置为49时,预测精度可以达到最高92.77%,测试后的内外回流比的拟合优度分别达到0.96316和0.90007;最后将训练后的最优值指导实践,检验效果。ELM优化后的内外回流比平均值分别为149.77%和72.12%,相比优化前200%和70%,分别降低25.12%和3.03%;污水处理厂依然可以实现稳定达标,并且总氮浓度平均值由6.08 mg/L下降到3.93 mg/L,降幅为33.22%。
Objective to improve the control level of wastewater treatment plant by intelligent optimization of internal reflux ratio of mixed liquid and external reflux ratio of sludge in AAO(anaerobic oxic)wastewater treatment process.The results show that the optimized reflux ratio not only reduces the dependence on artificial experience,but also effectively reduces the concentration of total nitrogen on the premise of ensuring the stable operation of sewage treatment plant.
作者
马兴冠
王志毅
贾澍
杨勇
MA Xingguan;WANG Zhiyi;JIA Shu;YANG Yong(School of Municipal and Environmental Engineering,Shenyang Jianzhu University,Shenyang 110168,China;Liaoning Ecological Environment Affairs Service Center,Shenyang 110161,China)
出处
《给水排水》
CSCD
北大核心
2021年第S02期479-483,共5页
Water & Wastewater Engineering
关键词
AAO
极限学习机
回流比
优化
AAO
Limit learning machine
Reflux ratio
Wisdom
