基于自组织递归小波神经网络的污水处理过程多变量控制

苏尹; 杨翠丽; 乔俊飞

doi:10.16383/j.aas.c220679

基于自组织递归小波神经网络的污水处理过程多变量控制

doi: 10.16383/j.aas.c220679

苏尹^{1, 2,},
杨翠丽^1,,
乔俊飞^1,

1.
北京工业大学信息学部, 计算智能与智能系统北京市重点实验室, 智慧环保北京实验室北京 100124
2.
嘉兴大学信息科学与工程学院嘉兴 314001

基金项目: 国家自然科学基金 (61890930-5, 62021003, 61973010), 国家重点研发计划(2021ZD0112302) 资助

详细信息

作者简介:
苏尹：嘉兴大学信息科学与工程学院讲师.2023年获得北京工业大学控制科学与工程博士学位. 主要研究方向为基于神经网络的城市污水处理过程预测及过程控制. E-mail: suy@zjxu.edu.cn

杨翠丽：北京工业大学信息学部副教授. 2008年获得中国石油大学(东营)工学学士学位, 2010年获得天津大学理学硕士学位, 2014年获得香港城市大学博士学位. 目前的研究方向包括计算智能, 污水处理过程的建模与控制. E-mail: clyang5@bjut.edu

乔俊飞：北京工业大学信息学部教授, 计算智能与智能系统北京市重点实验室主任. 分别于1992年和1995年在辽宁工业大学获得控制工程学士和硕士学位, 1998年在中国沈阳东北大学获得博士学位. 目前的研究方向包括神经网络、智能系统、自适应系统和过程控制. 本文通信作者. E-mail: adqiao@bjut.edu.cn

计量
- 文章访问数: 38
- HTML全文浏览量: 20
- 被引次数: 0
出版历程
- 网络出版日期: 2024-03-28

Multivariate Control of Wastewater Treatment Process Based on Self-organized Recurrent Wavelet Neural Network

SU Yin^{1, 2
,},
YANG Cui-Li^1
,,
QIAO Jun-Fei^1
,

1.
Beijing University of Technology, Faculty of Information Technology, Beijing Key Laboratory of Computational Intelligence and Intelligent System, Beijing Laboratory for Intelligent Environmental Protection, Beijing 100124
2.
Jiaxing University, College of Information Science and Engineering, Jiaxing 314001

Funds: Supported by National Science Foundation of China under Grants (61890930-5, 62021003, 61973010) and National Key Research and Development Program of China (2021ZD0112302)

More Information

Author Bio:
SU Yin　Lecturer in the college of Information Science and Engineering, Jiaxing University.She received her Ph.D degree in Control Science and Engineering from Beijing Institute of Technology in 2023. Her main research interests are neural network-based urban wastewater treatment process prediction and process control

YANG Cui-Li　Associate Professor in the Faculty of Information Technology, Beijing University of Technology. She received the B.E. degree from China University of Petroleum, Dongying, China, in 2008, the M.S. degree from Tianjin University, Tianjin, China in 2010, and the Ph.D. degree from City University of Hong Kong, Hong Kong, China, in 2014. Her current research interests include computational intelligence, modeling and control for wastewater treatment process

QIAO Jun-Fei　Professor in the Faculty of Information Technology, Beijing University of Technology, where he is also the Director of the Beijing Key Laboratory of Computational Intelligence and Intelligent Systems. He received the B.E. and M.E. degrees in control engineering from Liaoning Technical University, Fuxin, China, in 1992 and 1995, respectively, and the Ph.D. degree from Northeastern University, Shenyang, China, in 1998. His current research interests include neural networks, intelligent systems, self-adaptive systems, and process control. Corresponding author of this paper

摘要

摘要: 污水处理过程是一个包含多个生化反应的复杂过程, 具有非线性和动态特性. 因此, 实现污水处理过程的精准控制是一项挑战. 为了解决这个问题, 提出一种基于自组织递归小波神经网络的污水处理过程多变量控制. 首先, 针对污水处理过程的动态特性, 根据小波基的激活强度设计了一种自组织机制来动态调整递归小波神经网络控制器的结构, 提高控制的性能. 然后, 采用结合自适应学习率的在线学习算法, 实现控制器的参数学习. 此外, 通过李雅普诺夫稳定性定理证明了此控制器的稳定性. 最后, 采用基准仿真平台进行仿真验证, 实验结果表明, 此控制方法可以有效提高污水处理过程的控制绝对积分误差和平方误差积分的精度.
- 神经网络控制 /
- 污水处理过程 /
- 自组织机制 /
- 多变量控制
Abstract: The wastewater treatment process is a complex process containing multiple biochemical reactions with nonlinear and dynamic characteristics. Therefore, it is a challenge to achieve accurate control of the wastewater treatment process. To solve this problem, a multi-variable control of wastewater treatment process based on the self-organizing recurrent wavelet neural network is proposed. Firstly, to deal with the dynamicity of wastewater treatment process, according to the firing strength of the wavelet base, the self-organizing mechanism is designed to dynamically adjust the structure of the recurrent wavelet neural network controller to improve the control performance. Then, an online learning algorithm combined with adaptive learning rate is used to learn the parameters of controller. In addition, the stability of the controller is proved by the Lyapunov stability theorem. Finally, the benchmark simulation platform is used to conduct simulation.The experimental results show that this control method can effectively improve the integral of absolute error and integral of squared error of the wastewater treatment process.
- Neural network control /
- wastewater treatment process /
- self-organizing mechanism /
- multi-variable control

HTML全文

图 1 活性污泥法

Fig. 1 Activated sludge method

下载: 全尺寸图片幻灯片

图 2 控制框图

Fig. 2 Control block diagram

下载: 全尺寸图片幻灯片

图 3 SRWNN结构图

Fig. 3 The structure of SRWNN

下载: 全尺寸图片幻灯片

图 4 控制流程图

Fig. 4 The flow chart of control

下载: 全尺寸图片幻灯片

图 5 不同小波函数时DO控制结果

Fig. 5 Control results under different wavelet function

下载: 全尺寸图片幻灯片

图 6 SRWNN小波节点变化图

Fig. 6 Change of SRWNN wavelet node

下载: 全尺寸图片幻灯片

图 7 晴天工况下DO控制结果

Fig. 7 Control results of DO under dry condition

下载: 全尺寸图片幻灯片

图 8 晴天工况下NO控制结果

Fig. 8 Control results of NO under dry condition

下载: 全尺寸图片幻灯片

图 9 阴雨工况下DO控制结果

Fig. 9 Control results of DO under rain condition

下载: 全尺寸图片幻灯片

图 10 阴雨工况下NO控制结果

Fig. 10 Control results of NO under rain condition

下载: 全尺寸图片幻灯片

图 11 $K_{La5}$变化曲线

Fig. 11 The change of $K_{La5}$

下载: 全尺寸图片幻灯片

图 12 $Q_a$变化曲线

Fig. 12 The change of $Q_a$

下载: 全尺寸图片幻灯片

图 13 SRWNN小波节点变化图

Fig. 13 Change of SRWNN wavelet node

下载: 全尺寸图片幻灯片

图 14 晴天工况下DO控制结果

Fig. 14 Control results of DO under dry condition

下载: 全尺寸图片幻灯片

图 15 晴天工况下NO控制结果

Fig. 15 Control results of NO under dry condition

下载: 全尺寸图片幻灯片

图 16 阴雨工况下DO控制结果

Fig. 16 Control results of DO under rain condition

下载: 全尺寸图片幻灯片

图 17 阴雨工况下NO控制结果

Fig. 17 Control results of NO under rain condition

下载: 全尺寸图片幻灯片

图 18 $K_{La5}$变化曲线

Fig. 18 The change of $K_{La5}$

下载: 全尺寸图片幻灯片

图 19 $Q_a$变化曲线

Fig. 19 The change of $Q_a$

下载: 全尺寸图片幻灯片

表 1 不同控制方法在恒定设定值时的性能比较

Table 1 Performance comparison of different control methods at constant set-point

工况	控制器	No.	DO			NO
工况	控制器	No.	IAE	ISE	$ \text{Dev}^{\text{max}} $	IAE	ISE	$ \text{Dev}^{\text{max}} $
晴天	SRWNN	3	5.66×10^-4	1.63×10^-6	0.0087	0.0036	7.61×10^-5	0.0114
	RWNN	5	0.0017	3.26×10^-5	0.0526	0.002	3.06×10^-5	0.054
	NNOMC	10	0.0390*	5.31×10^-4*	0.0725*	0.0490*	7.18×10^-4*	0.1630*
	RARFNNC	4	0.0073*	1.61×10^-4*	0.0104*	0.0126*	2.83×10^-4*	0.1050*
	DRFNNC	6	0.0079*	1.82×10^-4*	0.0154*	0.0085*	3.25×10^-4*	0.0176*
阴雨	SRWNN	4	0.0041	1.75×10^-4	0.1042	0.0101	9.80×10^-4	0.1291
	RWNN	5	0.0051	2.21×10^-4	0.1434	0.0117	1.40×10^-3	0.2244
	PID	—	0.0016	1.90×10^-3	0.2038	0.0317	8.23×10^-3	0.3233
注: “$ \star $”表示原文中的结果, “—”表示无

下载: 导出CSV

表 2 不同控制方法在变化设定值时的性能比较

Table 2 Performance comparison of different control methods at changed set-point

工况	控制器	No.	DO			NO
工况	控制器	No.	IAE	ISE	$ \text{Dev}^{\text{max}} $	IAE	ISE	$ \text{Dev}^{\text{max}} $
晴天	SRWNN	3	0.0067	3.68×10^-6	0.0156	0.0061	1.64×10^-4	0.0067
	RWNN	5	0.0087	2.62×10^-4	0.1156	0.0126	2.30×10^-3	0.1116
	PID	-	0.0127	2.38×10^-3	0.1038	0.0271	4.90×10^-3	0.2184
阴雨	SRWNN	3	0.0047	1.10×10^-4	0.0538	0.0065	3.18×10^-4	0.1527
	RWNN	5	0.0069	1.92×10^-4	0.0644	0.0088	4.58×10^-4	0.1781
	RFNNC	—	0.0240*	2.40×10^-3*	0.0863	0.0260*	1.00×10^-3*	0.1881*
注: “$ \star $”表示原文中的结果, “—”表示无

下载: 导出CSV

参考文献(28)

[1]	Tang W Z, Pei Y S, Zheng H, Zhao Y, Shu L M, Zhang H. Twenty years of China's water pollution control: Experiences and challenges. Chemosphere, 2022, 295: 133875. doi: 10.1016/j.chemosphere.2022.133875
[2]	杨翠丽, 武战红, 韩红桂, 乔俊飞. 城市污水处理过程优化设定方法研究进展. 自动化学报, 2020, 46(10): 2092-2108. doi: 10.16383/j.aas.c200294 Yang C L, Wu Z H, Han H G, Qiao J F. Perspectives on optimal setting methods for municipal wastewater treatment processes. Acta Automatica Sinica, 2022, 46(10): 2092-2108. doi: 10.16383/j.aas.c200294
[3]	Lizarralde I, Fernandez-Arevalo T, Brouckaert C, Vanrolleghem P. A new general methodology for incorporating physico-chemical transformations into multi-phase wastewater treatment process models. Water Research, 2015, 74: 239-256. doi: 10.1016/j.watres.2015.01.031
[4]	潘南全. 基于前馈控制模型的生化反应池曝气控制优化. 工业控制计算机, 2015, 88(1): 58-59. doi: 10.3969/j.issn.1001-182X.2015.01.027 Pang N Q. Optimization aeration controI of reaction tank based on feed forward controI model. Industrial Control Computer, 2015, 88(1): 58-59. doi: 10.3969/j.issn.1001-182X.2015.01.027
[5]	Du S L, Yan Q S, Qiao J F. Event-triggered PID control for wastewater treatment plants. Journal of Water Process Engineering, 2020, 38: 101659. doi: 10.1016/j.jwpe.2020.101659
[6]	曾春霞, 董宗哲, 何涛. 模糊代数PID控制在污水处理溶解氧控制系统的应用. 化工自动化及仪表, 2021, 48(6): 528-534. doi: 10.3969/j.issn.1000-3932.2021.06.003 Zeng C X, Dong Z Z, He T. Application of fuzzy algebra PID control in dissolved oxygen control system of wastewater treatment. Control and Instruments in Chemical Industry, 2021, 48(6): 528-534. doi: 10.3969/j.issn.1000-3932.2021.06.003
[7]	刘锁清, 刘少红, 李军红, 彭伟娟. 基于模糊自整定PID串级控制的废水处理PH值控制. 自动化技术与应用, 2019, 38(2): 22-27. doi: 10.3969/j.issn.1003-7241.2019.02.006 Liu S Q, Liu S H, Li J H, Peng W J. Wastewater treatment PH value control based on fuzzy self-tuning PID cascade control. Techniques of Automation and Applications, 2019, 38(2): 22-27. doi: 10.3969/j.issn.1003-7241.2019.02.006
[8]	Hoang B L, Tien D N, Nguyen P H. Dissolved oxygen control of the activated sludge wastewater treatment process using Hedge Algebraic control. 2014 7th International Conference on Biomedical Engineering and Informatics, 2014: 827-832.
[9]	许进超, 杨翠丽, 乔俊飞, 马士杰. 基于自组织模糊神经网络溶解氧控制方法研究. 智能系统学报, 2018, 13(6): 905-912. doi: 10.11992/tis.201801019 Xu J C, Yang C L, Qiao J F, Ma S J. Dissolved oxygen concentration control method based on self-organizing fuzzy neural network. CAAI Transactions on Intelligent Systems, 2018, 13(6): 905-912. doi: 10.11992/tis.201801019
[10]	Wang D, Ha M M, Qiao J F. Data-driven iterative adaptive critic control toward an urban wastewater treatment plant. IEEE Transactions on Industrial Electronics, 2020, 68(8): 7362-7369.
[11]	Nawaz A, Arora A S, Yun C M, Lee J J, Lee M. Development of smart AnAmmOx system and its agile operation and decision support for pilot-scale WWTP. Soft Computing Techniques in Solid Waste and Wastewater Management, 2021: 423-7454.
[12]	韩红桂, 张璐, 卢薇, 乔俊飞. 城市污水处理过程动态多目标智能优化控制研究. 自动化学报, 2021, 47(3): 620-629. doi: 10.16383/j.aas.c190154 Han H G, Zhang L, Lu W, Qiao J F. Research on dynamic multiobjective intelligent optimal control for municipal wastewater treatment process. Acta Automatica Sinica, 2021, 47(3): 620-629. doi: 10.16383/j.aas.c190154
[13]	乔俊飞, 韩改堂, 周红标. 基于知识的污水生化处理过程智能优化方法. 自动化学报, 2017, 43(6): 1038-1046. doi: 10.16383/j.aas.2017.c170088 Qiao J F, Han G T, Zhou H B. Knowledge-based intelligent optimal control for wastewater biochemical treatment process. Acta Automatica Sinica, 2017, 43(6): 1038-1046. doi: 10.16383/j.aas.2017.c170088
[14]	Vega P, Revollar S, Francisco M, Martin J M. Integration of set point optimization techniques into nonlinear MPC for improving the operation of WWTPs. Computers and Chemical Engineering, 2014, 68(4): 78-95.
[15]	Han H G, Liu H X, Li J M, Qiao J F. Cooperative fuzzy-neural control for wastewater treatment process. IEEE Transactions on Industrial Informatics, 2021, 17(9): 5971-5981. doi: 10.1109/TII.2020.3034335
[16]	Ffm E S, Abuhasel K A. Adaptive nonlinear disturbance observer using double loop self-organizing recurrent wavelet-neural-network for two-axis motion control system. IEEE Transactions on Industry Applications, 2017, 54(1): 764-786.
[17]	王桐, 邱剑彬, 高会军. 随机非线性系统基于事件触发机制的自适应神经网络控制. 自动化学报, 2019, 45(1): 226-233. doi: 10.16383/j.aas.2018.c180404 Wang T, Qiu J B, Gao H J. Event-triggered adaptive neural network control for a class of stochastic nonlinear systems. Acta Automatica Sinica, 2019, 45(1): 226-233. doi: 10.16383/j.aas.2018.c180404
[18]	Liu Y J, Li J, Tong S, Chen L P. Neural network control-based adaptive learning design for nonlinear systems with full-state constraints. IEEE Transactions on Neural Networks and Learning Systems, 2016, 27(7): 1562-1571. doi: 10.1109/TNNLS.2015.2508926
[19]	Lee C H, Chang H H. Output recurrent wavelet neural network-based adaptive backstepping controller for a class of MIMO nonlinear non-affine uncertain systems. Neural Computing and Applications, 2014, 24(5): 1035-1045. doi: 10.1007/s00521-012-1326-2
[20]	Lin, Hong C. A novel hybrid recurrent wavelet neural network control of permanent magnet synchronous motor drive for electric scooter. Turkish Journal of Electrical Engineering and Computer Sciences, 2015, 22: 1056-1075.
[21]	张伟, 乔俊飞, 李凡军. 溶解氧浓度的直接自适应动态神经网络控制方法. 控制理论与应用, 2015, 32(1): 115-121. doi: 10.7641/CTA.2015.40311 Zhang W, Qiao J F, Li F J. Direct adaptive dynamic neural network control for dissolved oxygen concentration. Control Theory and Applications, 2015, 32(1): 115-121. doi: 10.7641/CTA.2015.40311
[22]	Fayez F M, Ei S, Khaled A A. Self-organizing recurrent fuzzy wavelet neural network-based mixed H2/H∞ adaptive tracking control for uncertain two-axis motion control system. 2015 IEEE Industry Applications Society Annual Meeting, 2015: 1-14.
[23]	Han H G, Zhang L, Hou Y, Qiao J F. Nonlinear model predictive control based on a self-organizing recurrent neural network. IEEE Transactions on Neural Networks and Learning Systems, 2016, 27(2): 402-415. doi: 10.1109/TNNLS.2015.2465174
[24]	Ku C C, Lee K Y. Diagonal recurrent neural networks for dynamics systems control. IEEE Transactions on Neural Network, 1995, 6(1): 144-156. doi: 10.1109/72.363441
[25]	韩广, 乔俊飞, 薄迎春. 溶解氧浓度的前馈神经网络建模控制方法. 控制理论与应用, 2013, 30(5): 585-591. doi: 10.7641/CTA.2013.20773 Han G, Qiao J F, Bo Y C. Feedforward neural network modeling and control for dissolved oxygen concentration. Control Theory and Applications, 2013, 30(5): 585-591. doi: 10.7641/CTA.2013.20773
[26]	Qiao J F, Han G T, Han H G, Chai W. Wastewater treatment control method based on a rule adaptive recurrent fuzzy neural network. International Journal of Intelligent Computing and Cybernetics, 2017, 2: 94-110.
[27]	Qiao J F, Han G T, Han H G, Yang C L, Chai W. Decoupling control for wastewater treatment process based on recurrent fuzzy neural network. Asian Journal of Control, 2018, 21(3): 1270-1280.
[28]	韩改堂, 乔俊飞, 韩红桂. 基于自适应递归模糊神经网络的污水处理控制. 控制理论与应用, 2016, 33(9): 1252-1258. doi: 10.7641/CTA.2016.50965 Han G T, Qiao J F, Han H G. Wastewater treatment control method based on adaptive recurrent fuzzy neural network. Control Theory and Applications, 2016, 33(9): 1252-1258. doi: 10.7641/CTA.2016.50965