基于领域自适应的城市污水处理运行过程多工况优化控制

扫码查看

原文链接

万方数据

中文摘要：城市污水处理运行过程的优化控制方法是提高其运行效率和改善其运行效果的关键.然而,由于城市污水处理过程进水负荷多变,运行过程存在多种工况且变化频繁,导致城市污水处理过程难以实现优化运行.因此,如何设计优化控制策略应对多种工况变化,保证出水总氮和出水总磷等水质指标达标,是城市污水处理过程亟需解决的挑战性问题.本文设计了基于领域自适应的城市污水处理运行过程多工况优化控制方法(multi-operating op-timization control with domain adaptive,MOOC-DA).首先,建立城市污水处理运行过程多工况优化目标模型,捕获运行能耗以及出水水质的时间序列特性,实现运行指标的精确预测.其次,设计基于多任务领域自适应粒子群的多工况优化设定方法,保证多工况运行出水水质达标.最后,设计基于多任务模糊神经网络的优化设定跟踪控制方法,实现城市污水处理过程多工况优化运行.为了验证所提出方法的有效性,基于活性污泥模型仿真平台将提出的MOOC-DA与其他优化控制方法进行对比实验.结果表明,该方法能够实现污水处理过程多工况的优化运行,保证出水总氮和出水总磷等水质指标达标.

外文标题：Multi-operating optimization control with domain adaptive for wastewater treatment operating process

外文摘要：The optimization control method for wastewater treatment process(WWTP)is the key to improving its operating efficiency and improving its operating effect.However,due to the variable influent pollutant load in WWTP,there are multiple operating conditions,and the changes are frequent,making it difficult for the optimal control system to achieve optimizing operation.Therefore,how to design an optimal control method to cope with changes in various operating conditions and ensure that water quality indicators such as total nitrogen and total phosphorus are up to standard is a challenging problem that needs to be solved in WWTP.This study designs a multi-operating optimization control method with domain adaptive(MOOC-DA)for WWTP.First,a multi-operating optimization objective model of WWTP is established to capture the time series characteristics of operation energy consumption and effluent quality,and achieve accurate prediction of operation indicators.Second,a multi-operating optimization setting method based on the multi-task domain adaptive is designed to ensure that the effluent quality meets the standard in multiple operating conditions.Finally,an optimal setting tracking control method based on multi-task fuzzy neural network is designed to realize multi-operating optimization operation of WWTP.At the same time,the proposed MOOC-DA is compared with other optimization control methods based on the activated sludge model No.2d simulation platform to verify the effectiveness.The results indicate that MOOC-DA can realize the optimal operation of multiple operating conditions in WWTP.

外文关键词：

wastewater treatment processmulti-operating optimization controlmulti-task optimizationdomain adaptive

作者：

韩红桂、白星、侯莹

展开 >

作者单位：

北京工业大学信息科学技术学院,北京 100124

计算智能与智能系统北京市重点实验室,北京 100124

数字社区教育部工程研究中心,北京 100124

关键词：

城市污水处理过程多工况优化控制多任务优化领域自适应

基金：

国家自然科学基金国家自然科学基金国家自然科学基金国家重点研发项目北京市杰出青年科学家计划北京市自然科学基金青年北京学者北京市教委研发项目

项目编号：

621253016202100361890930-52022YFB3305800-05BJJWZYJH01201910005020KZ202110005009037KM202210005027

出版年：

2024

DOI：

10.1360/SST-2023-0359

中国科学(技术科学)

中国科学院

中国科学(技术科学)

CSTPCD北大核心

影响因子：0.752

ISSN：1674-7259

年,卷(期)：2024.54(9)