基于改进集合经验模态分解和深度信念网络的出水总磷预测

doi:10.11949/0438-1157.20201365

摘要/Abstract

摘要：

出水总磷的准确预测对于城市污水处理厂的高效、稳定的运行至关重要。文中针对城市污水处理过程中出水总磷难以预测的问题，提出一种基于改进集合经验模态分解（modified ensemble empirical mode decomposition， MEEMD）和深度信念网络（deep belief network， DBN）的出水总磷预测方法。首先，设计一种MEEMD算法对城市污水处理过程出水总磷数据信号进行分解，获取多个本征模态函数（intrinsic mode function， IMF）组合；然后，建立一种基于模拟退火（simulated annealing， SA）算法的深度信念网络预测模型，通过优化的模型结构对分解后得到的每个IMF分量进行有效预测；最后，通过大气CO₂浓度预测和城市污水处理出水总磷预测验证了所提出方法的有效性。

关键词: 城市污水处理过程, 出水总磷, 集合经验模态分解, 深度信念网络

Abstract:

Accurate prediction of effluent total phosphorus is essential for the stable and efficient operation of urban wastewater treatment plants. Aiming at the problem that effluent total phosphorus is difficult to predict in urban wastewater treatment process, a prediction method of effluent total phosphorus based on modified ensemble empirical mode decomposition(MEEMD) and deep belief network (DBN) is proposed in this paper. First, a MEEMD algorithm is designed to decompose the effluent total phosphorus data signal of the effluent from urban wastewater treatment process. Then, establish a deep belief network prediction model based on simulated annealing (SA) algorithm, and effectively predict each IMF component obtained after decomposition through the optimized model structure. Finally, the effectiveness of the proposed method is verified by the prediction of atmospheric CO₂ concentration and the effluent total phosphorus in urban wastewater treatment.

Key words: urban sewage treatment process, effluent total phosphorus, ensemble empirical mode decomposition, deep belief network

中图分类号:

TP 183

王龙洋, 蒙西, 乔俊飞. 基于改进集合经验模态分解和深度信念网络的出水总磷预测[J]. 化工学报, 2021, 72(5): 2745-2753.

WANG Longyang, MENG Xi, QIAO Junfei. Prediction of effluent total phosphorus based on modified ensemble empirical mode decomposition and deep belief network[J]. CIESC Journal, 2021, 72(5): 2745-2753.

图/表 8

图1 深度信念网络模型结构图

Fig.1 Structure diagram of deep belief network model

图2 MEEMD-DBN-SA模型结构框图

Fig.2 Block diagram of MEEMD-DBN-SA model

图3 CO2浓度预测误差

Fig.3 Prediction error of CO2 concentration

图4 测试样本预测结果曲线

Fig.4 Prediction result curve of test sample

表1 不同算法性能比较

Table 1 Performance comparison of different algorithms

序号	预测算法	RMSE	APE	Time/s
1	BP	1.3000~6.6000		15.8
2	ALRDBN	1.1671		17.6
3	CDBN	1.1507		11.5
4	DBN	1.2830	0.0049	4.24
5	MEEMD-DBN	1.0051	0.0026	5.01
6	MEEMD-DBN-SA	0.3725	0.0008	10.56

图5 出水总磷浓度预测误差

Fig.5 Prediction error of total phosphorus concentration in effluent

图6 测试样本预测结果曲线

Fig.6 Prediction result curve of test sample

表2 不同算法性能比较

Table 2 Performance comparison of different algorithms

序号	预测算法	RMSE	APE	Time/s
1	ILM-SVDFNN	0.1061	0.1161	22.15
2	LM-SVDFNN	0.1078	0.1167	40.05
3	ILM-FNN	0.1296	0.1412	34.38
4	EBP-SVDFNN	0.1766	0.1690	85.76
5	EBP-FNN	0.1944	0.1840	114.37
6	DFNN	0.1763	0.1644	66.87
7	GDFNN	0.1539	0.1628	46.83
8	MPCA-NN	0.2668	—	68.93
9	RBF	0.0697	0.0969	12.19
10	GAP-RBF	0.1482	0.1843	100.00
11	IELM	0.0783	0.1105	91.17
12	APSO-RBF	0.0711	0.1168	222.63
13	DBN	0.0851	0.2428	0.70
14	MEEMD-DBN	0.0362	0.1564	36.29
15	MEEMD-DBN-SA	0.0269	0.1151	68.11

参考文献 27

28	Nicolson A, Paliwal K K. Deep learning for minimum mean-square error approaches to speech enhancement[J]. Speech Communication, 2019, 111: 44-55.
29	Chen L, Bian M Y, Luo Y G, et al. Real-time identification of the tyre–road friction coefficient using an unscented Kalman filter and mean-square-error-weighted fusion[J]. Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering, 2016, 230(6): 788-802.
1	Shao M L, Han X Q, Xie Z C, et al. Comparative study on macroinvertebrate communities along a reservoir cascade in Xiangxi River Basin[J]. Acta Ecologica Sinica, 2007, 27(12): 4963-4971.
2	Huang M Z, Ma Y W, Wan J Q, et al. A sensor-software based on a genetic algorithm-based neural fuzzy system for modeling and simulating a wastewater treatment process[J]. Applied Soft Computing, 2015, 27: 1-10.
3	杨琴, 谢淑云. BP神经网络在洞庭湖氨氮浓度预测中的应用[J]. 水资源与水工程学报, 2006, 17(1): 65-70.
	Yang Q, Xie S Y. Application of BP neural network into predicting NH₃-N concentration of Dongting Lake[J]. Journal of Water Resources and Water Engineering, 2006, 17(1): 65-70.
4	Hong S H, Lee M W, Lee D S, et al. Monitoring of sequencing batch reactor for nitrogen and phosphorus removal using neural networks[J]. Biochemical Engineering Journal, 2007, 35(3): 365-370.
5	Bagheri M, Mirbagheri S A, Ehteshami M, et al. Modeling of a sequencing batch reactor treating municipal wastewater using multi-layer perceptron and radial basis function artificial neural networks[J]. Process Safety and Environmental Protection, 2015, 93: 111-123.
6	Han H G, Lu W, Hou Y, et al. An adaptive-PSO-based self-organizing RBF neural network[J]. IEEE Transactions on Neural Networks and Learning Systems, 2018, 29(1): 104-117.
7	Li F J, Qiao J F, Han H G, et al. A self-organizing cascade neural network with random weights for nonlinear system modeling[J]. Applied Soft Computing, 2016, 42: 184-193.
8	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.
30	Liu S Q, Wu K, Zhou L J, et al. Modeling a dual-parallel silicon modulator for sinc-shaped nyquist pulse generation[J]. IEEE Journal of Selected Topics in Quantum Electronics, 2021, 27(3): 1-8.
31	Orekhov G, Luque J, Lerner Z F. Closing the loop on exoskeleton motor controllers: benefits of regression-based open-loop control[J]. IEEE Robotics and Automation Letters, 2020, 5(4): 6025-6032.
9	Wang Y X, Han H G, Guo M, et al. A self-organizing deep belief network based on information relevance strategy[J]. Neurocomputing, 2020, 396: 241-253.
10	Zou Q H, Xiong Q Y, Li Q D, et al. A water quality prediction method based on the multi-time scale bidirectional long short-term memory network[J]. Environmental Science and Pollution Research, 2020, 27(14): 16853-16864.
11	Srivastava N, Hinton G, Krizhevsky A, et al. Dropout: a simple way to prevent neural networks from overfitting[J]. Journal of Machine Learning Research, 2014, 15: 1929-1958.
12	Shen X, Tian X M, Liu T L, et al. Continuous dropout[J]. IEEE Transactions on Neural Networks and Learning Systems, 2018, 29(9): 3926-3937.
13	Hayashida T, Nishizaki I, Sekizaki S. Structural optimization of deep belief network by evolutionary computation methods including tabu search[J]. Transactions on Machine Learning and Artificial Intelligence, 2018, 6(1): 69.
14	Han M, Fan J C, Wang J. A dynamic feedforward neural network based on Gaussian particle swarm optimization and its application for predictive control[J]. IEEE Transactions on Neural Networks, 2011, 22(9): 1457-1468.
15	Li Y Y, Fang S K, Bai X Y, et al. Parallel design of sparse deep belief network with multi-objective optimization[J]. Information Sciences, 2020, 533: 24-42.
16	Bao L, Sun X Y, Chen Y, et al. Restricted Boltzmann machine-driven interactive estimation of distribution algorithm for personalized search[J]. Knowledge-Based Systems, 2020, 200: 106030.
17	Deng J F, Miriyala V P K, Zhu Z F, et al. Voltage-controlled spintronic stochastic neuron for restricted boltzmann machine with weight sparsity[J]. IEEE Electron Device Letters, 2020, 41(7): 1102-1105.
18	Hinton G E. Training products of experts by minimizing contrastive divergence[J]. Neural Computation, 2002, 14(8): 1771-1800.
19	Qin Y, Wang X, Zou J Q. The optimized deep belief networks with improved logistic sigmoid units and their application in fault diagnosis for planetary gearboxes of wind turbines[J]. IEEE Transactions on Industrial Electronics, 2019, 66(5): 3814-3824.
20	Hinton G E. Training products of experts by minimizing contrastive divergence[J]. Neural Computation, 2002, 14(8): 1771-1800.
21	Wang X P, Wang Y Q. A hybrid model of EMD and PSO-SVR for short-term load forecasting in residential quarters[J]. Mathematical Problems in Engineering, 2016, 2016(5): 1-10.
22	Cheng X X, Dong J, Peng Y, et al. A study of nonstationary wind effects on a full-scale large cooling tower using empirical mode decomposition[J]. Mathematical Problems in Engineering, 2017, 2017: 1-15.
23	罗曦, 王洪才, 李玉强. 基于CF-EEMD-LSSVR算法的铅冶炼系统温室气体排放的评估与预测[J]. 中南大学学报(自然科学版), 2018, 49(1): 15-21.
	Luo X, Wang H C, Li Y Q. Evaluation and prediction of greenhouse gas emission of lead smelting system based on CF-EEMD-LSSVR[J]. Journal of Central South University (Science and Technology), 2018, 49(1): 15-21.
24	Liu D, Sun K. Short-term PM_2.5 forecasting based on CEEMD-RF in five cities of China[J]. Environmental Science and Pollution Research, 2019, 26(32): 32790-32803.
25	刘厦, 刘石, 任婷. 基于SA-ELM的声学层析成像温度分布重建算法[J]. 化工学报, 2017, 68(6): 2434-2446.
	Liu S, Liu S, Ren T. SA-ELM based method for reconstructing temperature distribution in acoustic tomography measurement[J]. CIESC Journal, 2017, 68(6): 2434-2446.
26	张长兴, 王德水, 刘玉峰, 等. 模拟退火算法在岩土热物性参数确定中的应用[J]. 化工学报, 2015, 66(2): 545-552.
	Zhang C X, Wang D S, Liu Y F, et al. Application of simulated annealing algorithm for determining parameters of rock-soil thermal properties[J]. CIESC Journal, 2015, 66(2): 545-552.
27	Li G, Patel N A, Wang Y Z, et al. Fully actuated body-mounted robotic system for MRI-guided lower back pain injections: initial phantom and cadaver studies[J]. IEEE Robotics and Automation Letters, 2020, 5(4): 5245-5251.

[1]	徐圆, 张伟, 张明卿, 贺彦林. 基于FEEMD-AE与反馈极限学习机组合模型预测研究与应用[J]. 化工学报, 2018, 69(3): 1064-1070.
[2]	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.