化工学报 ›› 2019, Vol. 70 ›› Issue (5): 1868-1878.DOI: 10.11949/j.issn.0438-1157.20181530

• 过程系统工程 • 上一篇    下一篇

基于均匀分布NSGAII算法的污水处理多目标优化控制

李霏1,2(),杨翠丽1,2,李文静1,2,乔俊飞1,2()   

  1. 1. 北京工业大学信息学部,北京 100124
    2. 计算智能和智能系统北京市重点实验室,北京 100124
  • 收稿日期:2019-01-02 修回日期:2019-02-19 出版日期:2019-05-05 发布日期:2019-05-05
  • 通讯作者: 乔俊飞
  • 作者简介:<named-content content-type="corresp-name">李霏</named-content>(1985—),女,博士研究生,<email>lifeiaedu@hotmail.com</email>|乔俊飞(1968—),男,博士,教授,<email>junfeiq@bjut.edu.cn</email>
  • 基金资助:
    国家自然科学基金项目(61533002, 61603012, 61603009);北京市教委项目(KM201710005025);北京市博士后工作经费资助项目(2017ZZ-028);北京市朝阳区博士后工作经费资助项目(2017ZZ-01-07);北京市自然科学基金项目(4182007)

Optimal control of wastewater treatment process using NSGAII algorithm based on multi-objective uniform distribution

Fei LI1,2(),Cuili YANG1,2,Wenjing LI1,2,Junfei QIAO1,2()   

  1. 1. Faculty of Information Technology, Beijing University of Technology, Beijing 100124, China
    2. Beijing Key Laboratory of Computational Intelligence and Intelligent System, Beijing 100124, China
  • Received:2019-01-02 Revised:2019-02-19 Online:2019-05-05 Published:2019-05-05
  • Contact: Junfei QIAO

摘要:

针对污水处理过程控制中能耗过大、出水水质超标严重等问题,提出了一种基于均匀分布的NSGAII(non-dominated sorting genetic algorithm II based on uniform distribution, UDNSGAII)多目标优化智能控制系统。首先,该方法以污水处理能耗和出水水质作为优化目标,建立多目标优化模型。其次,为了获得溶解氧和硝态氮的优化设定值,提高Pareto解的性能,该算法将种群映射到目标函数对应的超平面, 并在该平面上进行聚类以增加解的多样性。此外,加入分布性判断模块和分布性加强模块提高解的分布性。最后,采用比例积分微分(proportional integral derivative, PID)控制器对溶解氧和硝态氮的优化设定值进行底层跟踪控制。为了验证该算法的有效性,采用国际基准的污水处理仿真平台(benchmark simulation model No.1, BSM1)来进行实验。结果显示,所提出的UDNSGAII多目标优化控制方法能够在满足出水水质达标的同时,有效地降低污水处理过程能耗。

关键词: 污水处理过程, 均匀分布, 遗传算法, 优化, 控制

Abstract:

Aiming at the problems of excessive energy consumption (EC) and exceeded seriously effluent quality (EQ) in wastewater treatment control process, an optimal control of wastewater treatment process using a multi-objective uniform distribution NSGAII algorithm (UDNSGAII) was proposed. Firstly, EC and EQ of wastewater treatment are regarded as optimization objectives, and the multi-objective optimal control model is established. Secondly, to obtain the optimal set values of dissolved oxygen (DO) and nitrate nitrogen (NO), and improve the performance of Pareto solution, the individuals which have been clustered are mapped to the hyperplane of the corresponding objective function, then, the diversity of population is increased. In addition, the distribution judgment module and distributed enhancement module are used to improve the distribution of solutions. Finally, PID controller as the bottom controller is used to track the optimal setting value of DO and NO. To test the effectiveness of the proposed algorithm, benchmark simulation model No.1 (BSM1) is used. The results show that the proposed UDNSGAII multi-objective optimization control method can effectively reduce EC of wastewater treatment process while meeting EQ standards.

Key words: wastewater treatment process, uniform distribution, genetic algorithm, optimal, control

中图分类号: