基于事件触发的污水处理过程水质智能预测研究

doi:10.11949/0438-1157.20241255

摘要/Abstract

摘要：

针对污水处理过程非平稳性、多工况性导致的水质难以高效、精准预测的问题，提出了一种事件触发的模糊神经网络（event-triggered fuzzy neural network，ETFNN）总磷预测模型设计方法，以事件的形式感知污水处理总磷状态演化过程的非平稳性和多工况性，进而实现对总磷状态高效、精准的跟踪和预测。首先，利用总磷历史数据对模糊神经网络（fuzzy neural network，FNN）进行训练，根据能够反映多工况切换的训练误差变化趋势来定义事件；其次，设计事件触发的模型参数更新策略，当不同事件发生时，模型会触发变步长的参数更新行为，即实现对污水处理运行非平稳性、多工况性的感知和识别；同时，通过构造等效Markov决策过程的性能势函数证明了ETFNN模型学习过程的收敛性；最后，将ETFNN作为软测量模型，用于实际污水处理过程出水总磷建模与预测，并进行了综合分析。实验结果表明，所提出的ETFNN软测量模型不仅能提高总磷预测精度，还能以事件的形式识别并跳过无效数据，进而降低预测模型的计算复杂度。

关键词: 污水处理, 水质预测, 事件触发, 模糊神经网络, Markov决策, 性能势函数

Abstract:

Aiming at the problem that the non-stationary and multi-working conditions of wastewater treatment process (WWTP) make it difficult to predict water quality efficiently and accurately, this paper proposes an event-triggered fuzzy neural network (ETFNN) model to predict total phosphorus (TP) of WWTP. This method can perceive the non-stationary and multi-working conditions during the evolution process of the TP state in the form of events, thereby achieving efficiently-accurately tracking and prediction. First, the fuzzy neural network (FNN) is trained using the historical data of total phosphorus, and events are defined according to the trend of training error changes that can reflect the switching of multiple operating conditions. Second, an event-triggered learning is designed to adaptively update the parameters of FNN, where the different learning steps will be triggered when some different events occur. This event-triggered learning can perceive and recognize the non-stationary and multiple operational conditions in WWTP. Meanwhile, the convergence analysis of the ETFNN model is given by analyzing the performance potential function of the equivalent Markov decision process. Finally, the ETFNN is considered as the soft-sensing model to predict TP of WWTP, and then a comprehensive analysis is given as well. Experimental results show that the proposed ETFNN-based soft-sensing model not only improves the accuracy of TP prediction, but also identifies and skips invalid data in the form of events, thereby reducing the computational complexity of the prediction model.

Key words: wastewater treatment, water-quality prediction, event-triggered, fuzzy neural network, Markov decision, performance potential function

中图分类号:

TP 183

李欣怡, 王功明, 王自鹏, 乔俊飞. 基于事件触发的污水处理过程水质智能预测研究[J]. 化工学报, 2025, 76(6): 2828-2837.

Xinyi LI, Gongming WANG, Zipeng WANG, Junfei QIAO. Research on intelligent prediction of water quality in sewage treatment process based on event triggering[J]. CIESC Journal, 2025, 76(6): 2828-2837.

图/表 8

图1 污水处理厂活性污泥法示意图

Fig.1 Schematic diagram of activated sludge in WWTP

图2 基于ETFNN的总磷软测量模型运行原理图

Fig.2 Operating schematic diagram of ETFNN-based total phosphorus soft-sensing model

图3 FNN的拓扑结构

Fig.3 Topology structure of FNN

图4 总磷训练结果及误差

Fig.4 Training results and errors of total phosphorus

图5 训练过程触发事件分布

Fig.5 Triggering instants of events in training process

图6 总磷预测结果及误差

Fig.6 Prediction results and errors of total phosphorus

表1 ETFNN软测量模型与其他模型在总磷浓度预测实验中的对比结果

Table 1 Comparison results of ETFNN soft-sensing model with the others in total phosphorus predictions

ETFNN

FNN

SOFNN

DBN

LSTMNN

RBFNN

BPNN

0.0415±0.0053

0.0861±0.0089

0.0646±0.0076

0.0598±0.0071

0.1094±0.0217

0.1208±0.0346

0.1572±0.0538

257±8 (48.60%±2.80%)↓

500

0.0436±0.0068

0.0977±0.0094

0.0703±0.0082

0.0697±0.0075

0.1273±0.0243

0.3745±0.0492

0.1783±0.6858

13.29±0.92

18.65±1.43

15.58±1.18

21.93±1.58

18.51±1.47

20.38±1.52

25.76±1.93

图7 总磷与其相关变量的关联性指数

Fig.7 Correlation index between TP and its related variables

参考文献 31

[1]	中华人民共和国国家发展和改革委员会. 关于推进污水资源化利用的指导意见[R]. (2021-01-11)[2024-11-06].https://www.ndrc.gov.cn/xwdt/tzgg/202101/t20210111_1264795.html.
	National Development and Reform Commission. Duidances on promoting the utilization of polluted water resources[R]. (2021-01-11)[2024-11-06]. https://www.ndrc.gov.cn/xwdt/tzgg/202101//t20210111_1264795.html.
[2]	徐敏, 张涛, 王东, 等. 中国水污染防治40年回顾与展望[J]. 中国环境管理, 2019, 11(3): 65-71.
	Xu M, Zhang T, Wang D, et al. Review and prospect of water pollution prevention and control of China in the forty years of reform and opening-up[J]. Chinese Journal of Environmental Management, 2019, 11(3): 65-71.
[3]	蔡婕萍, 左高山, 许和连. 虚拟水: 国际水资源管理与启示[J]. 生态经济, 2019, 35(10): 200-206.
	Cai J P, Zuo G S, Xu H L. Virtual water: the ana-lysis and enlightment of international water resources management[J]. Ecological Economy, 2019, 35(10): 200-206.
[4]	王功明, 李文静, 乔俊飞. 基于PLSR自适应深度信念网络的出水总磷预测[J]. 化工学报, 2017, 68(5): 1987-1997.
	Wang G M, Li W J, Qiao J F. Prediction of effluent total phosphorus using PLSR-based adaptive deep belief network[J]. CIESC Journal, 2017, 68(5): 1987-1997.
[5]	张帅, 周平. 污水处理过程递推双线性子空间建模及无模型自适应控制[J]. 自动化学报, 2022, 48(7): 1747-1759.
	Zhang S, Zhou P. Recursive dual-line subspace modeling and model-free adaptive control in sewage treatment process[J]. Acta Automatica Sinica, 2022, 48(7): 1747-1759.
[6]	Wang G M, Jia Q S, Qiao J F, et al. A sparse deep belief network with efficient fuzzy learning framework[J]. Neural Networks, 2020, 121: 430-440.
[7]	郭鑫, 李文静, 乔俊飞. 基于自组织模块化神经网络的污水处理过程出水参数预测[J]. 化工学报, 2024, 75(9): 3242-3254.
	Guo X, Li W J, Qiao J F. Prediction of effluent parameters in wastewater treatment process using self-organizing modular neural network[J]. CIESC Journal, 2024, 75(9): 3242-3254.
[8]	王康成, 尚超, 柯文思, 等. 化工过程深度神经网络软测量的结构与参数自动调整方法[J]. 化工学报, 2018, 69(3): 900-906.
	Wang K C, Shang C, Ke W S, et al. Automatic structure and parameters tuning method for deep neural network soft sensor in chemical industries[J]. CIESC Journal, 2018, 69(3): 900-906.
[9]	Wang G M, Jia Q S, Zhou M C, et al. Artificial neural networks for water quality soft-sensing in wastewater treatment: a review[J]. Artificial Intelligence Review, 2022, 55(1): 565-587.
[10]	Foschi J, Turolla A, Antonelli M. Soft sensor predictor of E.coli concentration based on conventional monitoring parameters for wastewater disinfection control[J]. Water Research, 2021, 191: 116806.
[11]	韩红桂, 陈治远, 乔俊飞, 等. 基于区间二型模糊神经网络的出水氨氮软测量[J]. 化工学报, 2017, 68(3): 1032-1040.
	Han H G, Chen Z Y, Qiao J F, et al. Soft-sensor method for effluent ammonia nitrogen based on interval type-2 fuzzy neural networks[J]. CIESC Journal, 2017, 68(3): 1032-1040.
[12]	廉小亲, 王俐伟, 安飒, 等. 基于SOM-RBF神经网络的COD软测量方法[J]. 化工学报, 2019, 70(9): 3465-3472.
	Lian X Q, Wang L W, An S, et al. On soft sensor of chemical oxygen demand by SOM-RBF neural network[J]. CIESC Journal, 2019, 70(9): 3465-3472.
[13]	闻超垚, 周平. 污水处理过程出水水质稀疏鲁棒建模[J]. 自动化学报, 2022, 48(6): 1469-1481.
	Wen C Y, Zhou P. Sparse robust modeling of effluent quality indices in wastewater treatment process[J]. Acta Automatica Sinica, 2022, 48(6): 1469-1481.
[14]	张瑞垚, 周平. 基于鲁棒加权模糊聚类的污水处理过程监测方法[J]. 自动化学报, 2022, 48(9): 2198-2211.
	Zhang R Y, Zhou P. Robust weighted fuzzy clustering for sewage treatment process monitoring[J]. Acta Automatica Sinica, 2022, 48(9): 2198-2211.
[15]	李文静, 李萌, 乔俊飞. 基于互信息和自组织RBF神经网络的出水BOD软测量方法[J]. 化工学报, 2019, 70(2): 687-695.
	Li W J, Li M, Qiao J F. Effluent BOD soft measurement based on mutual information and self-organizing RBF neural network[J]. CIESC Journal, 2019, 70(2): 687-695.
[16]	曹跃, 陈志文, 袁小锋, 等. 部分子块通讯的分布式PCA厂级工业过程监测方法[J]. 控制与决策, 2020, 35(6): 1281-1290.
	Cao Y, Chen Z W, Yuan X F, et al. Distributed PCA for plant-wide processes monitoring with partial block communication[J]. Control and Decision, 2020, 35(6): 1281-1290.
[17]	Heo S, Nam K, Loy-Benitez J, et al. Data-driven hybrid model for forecasting wastewater influent loads based on multimodal and ensemble deep learning[J]. IEEE Transactions on Industrial Informatics, 2021, 17(10): 6925-6934.
[18]	钱锋, 杜文莉, 钟伟民, 等. 石油和化工行业智能优化制造若干问题及挑战[J]. 自动化学报, 2017, 43(6): 893-901.
	Qian F, Du W L, Zhong W M, et al. Problems and challenges of smart optimization manufacturing in petrochemical industries[J]. Acta Automatica Sinica, 2017, 43(6): 893-901.
[19]	韩红桂, 张琳琳, 伍小龙, 等. 数据和知识驱动的城市污水处理过程多目标优化控制[J]. 自动化学报, 2021, 47(11): 2538-2546.
	Han H G, Zhang L L, Wu X L, et al. Data-knowledge driven multiobjective optimal control for municipal wastewater treatment process[J]. Acta Automatica Sinica, 2021, 47(11): 2538-2546.
[20]	佟素娟, 薛同来. 基于PSO-ACO算法的再生水厂出水总磷预测模型研究[J]. 现代盐化工, 2023, 50(4): 35-37.
	Tong S J, Xue T L. Study on prediction model of total phosphorus in effluent of reclaimed water plant based on PSO-ACO algorithm[J]. Modern Salt Chemical Industry, 2023, 50(4): 35-37.
[21]	Wang K, Yuan X F, Chen J, et al. Supervised and semi-supervised probabilistic learning with deep neural networks for concurrent process-quality monitoring[J]. Neural Networks, 2021, 136: 54-62.
[22]	孙铭, 魏守科, 王莹洁, 等. 基于小波分解的LSTM水质预测模型[J]. 计算机系统应用, 2020, 29(12): 55-63.
	Sun M, Wei S K, Wang Y J, et al. Prediction model of water quality based on wavelet decomposition and LSTM[J]. Computer Systems & Applications, 2020, 29(12): 55-63.
[23]	Zhou H B, Qiao J F. Soft sensing of effluent ammonia nitrogen using rule automatic formation-based adaptive fuzzy neural network[J]. Desalination and Water Treatment, 2019, 140: 132-142.
[24]	Wang G M, Qiao J F. An efficient self-organizing deep fuzzy neural network for nonlinear system modeling[J]. IEEE Transaction on Fuzzy Systems, 2022, 30(7): 2170-2182.
[25]	Du S L, Zhang Q D, Han H G, et al. Event-triggered model predictive control of wastewater treatment plants[J]. Journal of Water Process Engineering, 2022, 47: 102765.
[26]	Liu Z, Han H G, Yang H Y, et al. Knowledge-aided and data-driven fuzzy decision making for sludge bulking[J]. IEEE Transactions on Fuzzy Systems, 2023, 31(4): 1189-1201.
[27]	Wang G M, Chen H, Han H G, et al. Predicting water quality with nonstationarity: event-triggered deep fuzzy neural network[J]. IEEE Transactions on Fuzzy Systems, 2024, 32(5): 2690-2699.
[28]	Wu Z J, Jia Q S, Guan X H. Optimal control of multiroom HVAC system: an event-based approach[J]. IEEE Transactions on Control Systems Technology, 2016, 24(2): 662-669.
[29]	Cao X R. Stochastic learning and optimization—A sensitivity-based approach[J]. Annual Reviews in Control, 2009, 33(1):11-24.
[30]	Wang G M, Chen H, Jiang S L, et al. Neurodynamics-driven prediction model for state evolution of coastal water quality[J]. IEEE Transactions on Instrumentation and Measurement, 2024, 73: 2519409.
[31]	Bi J, Lin Y Z, Dong Q X, et al. Large-scale water quality prediction with integrated deep neural network[J]. Information Sciences, 2021, 571: 191-205.