基于PLSR自适应深度信念网络的出水总磷预测

doi:10.11949/j.issn.0438-1157.20161826

化工学报 ›› 2017, Vol. 68 ›› Issue (5): 1987-1997.DOI: 10.11949/j.issn.0438-1157.20161826

基于PLSR自适应深度信念网络的出水总磷预测

王功明^1,2, 李文静^1,2, 乔俊飞^1,2

1 北京工业大学信息学部, 北京 100124;
2 计算智能与智能系统北京市重点实验室, 北京 100124

收稿日期:2016-12-29 修回日期:2017-02-14 出版日期:2017-05-05 发布日期:2017-05-05
通讯作者: 王功明
基金资助:
国家杰出青年科学基金项目（61225016）；国家自然科学基金项目（61533002，61603009）；中国博士后科学基金项目（2015M570910）；朝阳区博士后科研基金项目（2015ZZ-6）；北京工业大学基础研究基金项目（002000514315501）。

Prediction of effluent total phosphorus using PLSR-based adaptive deep belief network

WANG Gongming^1,2, LI Wenjing^1,2, QIAO Junfei^1,2

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:2016-12-29 Revised:2017-02-14 Online:2017-05-05 Published:2017-05-05
Supported by:
supported by the National Science Foundation for Distinguished Young Scholars of China (61225016), the National Natural Science Foundation of China (61533002, 61603009), the China Postdoctoral Science Foundation (2015M570910), the Chaoyang District Postdoctoral Research Foundation (2015ZZ-6), the Basic Research Foundation Project of Beijing University of Technology (002000514315501).

摘要/Abstract

摘要：

针对污水处理过程出水总磷预测问题存在的强非线性、大时变等特征，提出了一种基于偏最小二乘回归自适应深度信念网络（partial least square regression adaptive deep belief network，PLSR-ADBN）的出水总磷预测方法。PLSR-ADBN是基于深度学习模型DBN的一种改进型建模方法。首先，将自适应学习率引入到DBN的无监督预训练（pre-training）阶段，来提高网络收敛速度。其次，利用PLSR方法取代传统DBN中基于梯度的逐层权值精调（fine-tuning）方法，来提高网络预测精度。同时，通过构造李雅普诺夫函数证明了PLSR-ADBN学习过程的收敛性。最后，将PLSR-ADBN用于实际污水处理过程出水总磷预测中。实验结果表明所提出的PLSR-ADBN收敛速度快且预测精度高，能够满足实际污水处理过程对出水总磷监测精度和运行效率的要求。

关键词: 出水总磷, PLSR, 自适应学习率, 深度学习, 深度信念网络

Abstract:

Considered high nonlinearity and large transient variation, a PLSR-adaptive deep belief network (PLSR-ADBN) was proposed for prediction of total phosphorus (TP) in effluent of wastewater treatment process (WWTP). The PLSR-ADBN was an improved DBN, a deep learning model. First, an adaptive learning rate was introduced into the unsupervised pre-training stage of DBN so as to accelerate convergence rate. Secondly, PLSR was used to replace gradient fine-tuning method in conventional DBN for improving prediction accuracy. Meanwhile, a Lyapunov function was constructed to prove convergence of the PLSR-ADBN learning process. Finally, the proposed PLSR-ADBN was applied to an actual TP prediction in WWTP. The experimental results show that the method has a fast convergence rate and a high prediction accuracy, which can meet the demands for TP detection accuracy and WWTP operating efficiency.

Key words: total phosphorus in effluent, PLSR, adaptive learning rate, deep learning, deep belief network

中图分类号:

TP183

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 Gongming, LI Wenjing, QIAO Junfei. Prediction of effluent total phosphorus using PLSR-based adaptive deep belief network[J]. CIESC Journal, 2017, 68(5): 1987-1997.

参考文献 34

[1]	SHAO M, HAN X, XIE Z, 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]	LIU L, LIU D, JOHNSON D M, et al. Effects of vertical mixing on phytoplankton blooms in Xiangxi Bay of Three Gorges Reservoir: implications for management[J]. Water Research, 2012, 46(7): 2121-2130.
[3]	JIANG L, LIANG B, XUE Q, et al. Characterization of phosphorus leaching from phosphate waste rock in the Xiangxi River watershed, Three Gorges Reservoir, China[J]. Chemosphere, 2016, 150: 130-138.
[4]	DUPAS R, DELMAS M, DORIOZ J M, et al. Assessing the impact of agricultural pressures on N and P loads and eutrophication risk[J]. Ecological Indicators, 2015, 48: 396-407.
[5]	SUNOHARA M D, GOTTSCHALL N, CRAIOVAN E, et al. Controlling tile drainage during the growing season in Eastern Canada to reduce nitrogen, phosphorus, and bacteria loading to surface water[J]. Agricultural Water Management, 2016, 178: 159-170.
[6]	JIN L, ZHANG G, TIAN H. Current state of sewage treatment in China[J]. Water Research, 2014, 66: 85-98.
[7]	QIAO J F, LI W, HAN H G. Soft computing of biochemical oxygen demand using an improved T-S fuzzy neural network[J]. Chinese Journal of Chemical Engineering, 2014, 22(11): 1254-1259.
[8]	SHAO W, TIAN X, WANG P. Local partial least squares based online soft sensing method for multi-output processes with adaptive process states division[J]. Chinese Journal of Chemical Engineering, 2014, 22(7): 828-836.
[9]	许少鹏, 韩红桂, 乔俊飞. 基于模糊递归神经网络的污泥容积指数预测模型[J]. 化工学报, 2013, 64 (12): 4550-4556. XU S P, HAN H G, QIAO J F. Prediction of activated sludge bulking based on recurrent fuzzy neural network[J]. CIESC Journal, 2013, 64(12): 4550-4556.
[10]	QIAO J F, HAN G, HAN H G. Neural network on-line modeling and controlling method for multi-variable control of wastewater treatment processes[J]. Asian Journal of Control, 2014, 16(4): 1213-1223.
[11]	AKRATOS C S, PAPASPYROS J N E, TSIHRINTZIS V A. Artificial neural network use in ortho-phosphate and total phosphorus removal prediction in horizontal subsurface flow constructed wetlands[J]. Biosystems Engineering, 2009, 102(2): 190-201.
[12]	LI F J, QIAO J, 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.
[13]	CYBENKO G. Approximation by superpositions of a sigmoidal function[J]. Mathematics of Control, Signals and Systems, 1989, 2(4): 303-314.
[14]	TAMURA S, TATEISHI M. Capabilities of a four-layered feedforward neural network: four layers versus three[J]. IEEE Transactions on Neural Networks, 1997, 8(2): 251-255.
[15]	HINTON G E, SALAKHUTDINOV R R. Reducing the dimensionality of data with neural networks[J]. Science, 2006, 313(5786): 504-507.
[16]	HINTON G E, OSINDERO S, TEH Y W. A fast learning algorithm for deep belief nets[J]. Neural Computation, 2006, 18(7): 1527-1554.
[17]	ROSA E D L, YU W. Randomized algorithms for nonlinear system identification with deep learning modification[J]. Information Sciences, 2016, 364: 197-212.
[18]	LOPES N, RIBEIRO B. Towards adaptive learning with improved convergence of deep belief networks on graphics processing units[J]. Pattern Recognition, 2014, 47(1): 114-127.
[19]	LECUN Y, BENGIO Y, HINTON G. Deep learning[J]. Nature, 2015, 521(7553): 436-444.
[20]	ROWINSKI P M, PIOTROWSKI A. Estimation of parameters of the transient storage model by means of multi-layer perceptron neural networks/estimation des paramètres du modèle de transport TSM au moyen de réseaux de neurones perceptrons multi-couches[J]. Hydrological Sciences Journal, 2008, 53(1): 165-178.
[21]	PIOTROWSKI A P, NAPIORKOESKI J J. Optimizing neural networks for river flow forecasting-evolutionary computation methods versus the Levenberg-Marquardt approach[J]. Journal of Hydrology, 2011, 407(1): 12-27.
[22]	ZUR R M, JIANG Y, PESCE L L, et al. Noise injection for training artificial neural networks: a comparison with weight decay and early stopping[J]. Medical Physics, 2009, 36(10): 4810-4818.
[23]	JAGANNATHAN S, LEWIS F L. Identification of nonlinear dynamical systems using multilayered neural networks[J]. Automatica, 1996, 32(12): 1707-1712.
[24]	HE Y L, GENG Z Q, XU Y, et al. A robust hybrid model integrating enhanced inputs based extreme learning machine with PLSR (PLSR-EIELM) and its application to intelligent measurement[J]. ISA Transactions, 2015, 58: 533-542.
[25]	JANIK L J, FORRESTER S T, RAWSON A. The prediction of soil chemical and physical properties from mid-infrared spectroscopy and combined partial least-squares regression and neural networks (PLS-NN) analysis[J]. Chemometrics and Intelligent Laboratory Systems, 2009, 97(2): 179-188.
[26]	MOHAMED A, DAHL G E, HINTON G E. Acoustic modeling using deep belief networks[J]. IEEE Transactions on Audio, Speech, and Language Processing, 2012, 20(1): 14-22.
[27]	HINTON G E. Training products of experts by minimizing contrastive divergence[J]. Neural Computation, 2002, 14(8): 1771-1800.
[28]	LEROUX N, BENGIO Y. Representational power of restricted boltzmann machines and deep belief networks[J]. Neural Computation, 2008, 20(6): 1631-1649.
[29]	BENGIO Y, DELALLEAU O. Justifying and generalizing contrastive divergence[J]. Neural Computation, 2009, 21(6): 1601-1621.
[30]	JI N N, ZHANG J S, ZHANG C X. A sparse-response deep belief network based on rate distortion theory[J]. Pattern Recognition, 2014, 47(9): 3179-3191.
[31]	MENG J E, WU S Q. A fast learning algorithm for parsimonious fuzzy neural systems[J]. Fuzzy Sets and Systems, 2002, 126(3): 337-351.
[32]	KIM M H, KIM Y S, PRABU A A, et al. A systematic approach to data-driven modeling and soft sensing in a full-scale plant[J]. Water Science and Technology, 2009, 60(2): 363-370.
[33]	WU S, ER M J, GAO Y. A fast approach for automatic generation of fuzzy rules by generalized dynamic fuzzy neural networks[J]. IEEE Transactions on Fuzzy Systems, 2001, 9(4): 578-594.
[34]	乔俊飞, 潘广源, 韩红桂. 一种连续型深度信念网的设计与应用[J]. 自动化学报, 2015, 41(12): 2138-2146. QIAO J F, PAN G Y, HAN H G. Design and application of continuous deep belief network[J]. Acta Automatica Sinica, 2015, 41(12): 2138-2146.

基于PLSR自适应深度信念网络的出水总磷预测

Prediction of effluent total phosphorus using PLSR-based adaptive deep belief network

PDF (PC)

可视化

被引次数

摘要/Abstract

引用本文

使用本文

参考文献 34

相关文章 14

编辑推荐

Metrics

本文评价

[1]	温凯杰, 郭力, 夏诏杰, 陈建华. 一种耦合CFD与深度学习的气固快速模拟方法[J]. 化工学报, 2023, 74(9): 3775-3785.
[2]	齐书平, 王文龙, 张磊, 都健. 基于深度学习的金属离子-有机配体配位稳定常数的预测[J]. 化工学报, 2022, 73(12): 5461-5468.
[3]	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.
[4]	谢昊源, 黄群星, 林晓青, 李晓东, 严建华. 基于图像深度学习的垃圾热值预测研究[J]. 化工学报, 2021, 72(5): 2773-2782.
[5]	陈忠圣, 朱梅玉, 贺彦林, 徐圆, 朱群雄. 基于分位数回归CGAN的虚拟样本生成方法及其过程建模应用[J]. 化工学报, 2021, 72(3): 1529-1538.
[6]	于程远, 吴金奎, 周利, 吉旭, 戴一阳, 党亚固. 基于深度学习预测有机光伏电池能量转换效率[J]. 化工学报, 2021, 72(3): 1487-1495.
[7]	王晓慧, 王延江, 邓晓刚, 张政. 基于加权深度支持向量数据描述的工业过程故障检测[J]. 化工学报, 2021, 72(11): 5707-5716.
[8]	尹林子, 关羽吟, 蒋朝辉, 许雪梅. 基于k-means++的高炉铁水硅含量数据优选方法[J]. 化工学报, 2020, 71(8): 3661-3670.
[9]	周红标, 张钰, 柏小颖, 刘保连, 赵环宇. 基于自适应模糊神经网络的非线性系统模型预测控制[J]. 化工学报, 2020, 71(7): 3201-3212.
[10]	田璐瑶, 王梓豪, 粟杨, 文华强, 申威峰. 基于深度学习的溶剂定量构效关系建模研究进展[J]. 化工学报, 2020, 71(10): 4462-4472.
[11]	黄正梁, 王超, 李少硕, 杨遥, 孙婧元, 王靖岱, 阳永荣. 基于深度学习的气液固三相反应器图像分析方法及应用[J]. 化工学报, 2020, 71(1): 274-282.
[12]	顾恒昌, 牟鹏, 李建伟. 基于交叉迭代BLSTM网络的乙烯裂解炉建模[J]. 化工学报, 2019, 70(2): 548-555.
[13]	董顺, 李益国, 孙栓柱, 刘西陲, 沈炯. 基于状态空间主成分分析网络的故障检测方法[J]. 化工学报, 2018, 69(8): 3528-3536.
[14]	王康成, 尚超, 柯文思, 江永亨, 黄德先. 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.