A self-adaptive multi-output soft sensor modeling based on deep neural network

doi:10.11949/j.issn.0438-1157.20171624

Abstract

Abstract:

In wastewater treatment process (WWTP), existence of several important but difficult-to-measure process variables hinders not only monitoring of the production process but also adjustment or optimization of process control strategies. Even though soft-sensor models are reasonably constructed, it will still suffer degradation problem and high maintenance cost in real-time operation. Additionally, selection of proper secondary variables directly affects subsequent modeling. Therefore, a self-adaptive multi-output soft sensor model based on deep neural network was proposed for simultaneous online prediction of multiple target variables in wastewater treatment. Deep neural network was constructed from a stacked auto-encoder, which had satisfactory performance of online prediction under extremely complex scenarios. In order to overcome degradation problem and select proper secondary variables, time difference modeling and VIP (variable importance in projection) methods were added. Finally, validation on a true WWTP process shows that the proposed soft-sensor model has good performance on multiple output prediction and satisfactory prediction on single target.

Key words: wastewater, soft sensor, neural network, multi-output, prediction, time difference modeling, VIP

摘要：

在污水处理运行过程中，多个重要的难测过程变量的存在，不仅妨碍了生产过程的监控，而且阻碍了过程控制策略的调整或优化。即使软测量模型得到合理的构建，在投入运行后仍然遭受性能的退化和同时带来的高昂的维护成本。此外，合适辅助变量的选取直接影响后续建模的效果。因此，文中提出了一种基于深层神经网络的多输出自适应软测量模型，用于污水处理过程中多个目标变量的同步在线预测。其中，深层神经网络基于一种栈式自编码而构建，在极端复杂场景下具有优异的在线预测性能；并在建模中引入时差建模和变量重要性投影（VIP）这两种算法，以应对性能退化问题和实现辅助变量的精选。最后，通过一个实际案例对所提出模型进行验证。结果表明，所提出的软测量模型不仅具有较好的多输出预测性能，且在单目标预测结果上也有不错的表现。

关键词: 污水, 软测量, 神经网络, 多输出, 预测, 时差建模, 变量重要性投影

CLC Number:

TP277

QIU Yu, LIU Yiqi, WU Jing, HUANG Daoping. A self-adaptive multi-output soft sensor modeling based on deep neural network[J]. CIESC Journal, 2018, 69(7): 3101-3113.

邱禹, 刘乙奇, 吴菁, 黄道平. 基于深层神经网络的多输出自适应软测量建模[J]. 化工学报, 2018, 69(7): 3101-3113.

References 30

[1]	黄道平, 刘乙奇, 李艳. 软测量在污水处理过程中的研究与应用[J]. 化工学报, 2011, 62(1):1-9. HUANG D P, LIU Y Q, LI Y. Soft sensor research and its application in wastewater treatment[J]. CIESC Journal, 2011, 62(1):1-9.
[2]	曹鹏飞, 罗雄麟. 化工过程软测量建模方法研究进展[J]. 化工学报, 2013, 64(3):788-800. CAO P F, LUO X L. Modeling of soft sensor for chemical process[J]. CIESC Journal, 2013, 64(3):788-800.
[3]	KANO M, NAKAGAWA Y. Data-based process monitoring, process control, and quality improvement:recent developments and applications in steel industry[J]. Computers & Chemical Engineering, 2008, 32(1/2):12-24.
[4]	KADLEC P, GABRYS B, STRANDT S. Data-driven soft sensors in the process industry[J]. Computers & Chemical Engineering, 2009, 33(4):795-814.
[5]	BORCHANI H, VARANDO G, BIELZA C. A survey on multi-output regression[J]. Wiley Interdisciplinary Reviews Data Mining & Knowledge Discovery, 2015, 5(5):216-233.
[6]	ZHONG W, YU J. MIMO soft sensors for estimating product quality with on-line correction[J]. Chemical Engineering Research & Design, 2000, 78(4):612-620.
[7]	KOCEV D, D?EROSKI S, WHITE M D, et al. Using single-and multi target regression trees and ensembles to model a compound index of vegetation condition[J]. Ecological Modelling, 2009, 220(8):1159-1168.
[8]	BREIMAN L, FRIEDMAN J H. Predicting multivariate responses in multiple linear regression[J]. Journal of the Royal Statistical Society, 1997, 59(1):3-54.
[9]	Similä T, TIKKA J. Input selection and shrinkage in multiresponse linear regression[J]. Computational Statistics & Data Analysis, 2007, 52(1):406-422.
[10]	FU Y, SU H, CHU J. MIMO soft-sensor model of nutrient content for compound fertilizer based on hybrid modeling technique 1[J]. Chinese Journal of Chemical Engineering, 2007, 15(4):554-559.
[11]	郭民, 祝曙光, 韩红桂. 基于模糊神经网络的出水总磷和氨氮软测量方法研究[J]. 计算机与应用化学, 2017, 34(1):79-84. GUO M, ZHU S G, HAN H G. Soft-sensor method for total phosphorus and ammonia nitrogen based on fuzzy neural networks[J]. Computers and Applied Chemistry, 2017, 34(1):79-84.
[12]	QIU Y, LIU Y, HUANG D. Date-driven soft-sensor design for biological wastewater treatment using deep neural networks and genetic algorithms[J]. Journal of Chemical Engineering of Japan, 2016, 49(10):925-936.
[13]	KADLEC P, GRBI? R, GABRYS B. Review of adaptation mechanisms for data-driven soft sensors[J]. Computers & Chemical Engineering, 2011, 35(1):1-24.
[14]	JIANG J H, BERRY R J, SIESLER H W, et al. Wavelength interval selection in multicomponent spectral analysis by moving window partial least-squares regression with applications to mid-infrared and near-infrared spectroscopic data[J]. Analytical Chemistry, 2002, 74(14):3555-3565.
[15]	DAYAL B S, MACGREGOR J F. Recursive exponentially weighted PLS and its applications to adaptive control and prediction[J]. Journal of Process Control, 1997, 7(3):169-179.
[16]	QIN S J. Recursive PLS algorithms for adaptive data modeling[J]. Computers & Chemical Engineering, 1998, 22(4/5):503-514.
[17]	CHENG C, CHIU M S. A new data-based methodology for nonlinear process modeling[J]. Chemical Engineering Science, 2004, 59(13):2801-2810.
[18]	FUJIWARA K, KANO M, HASEBE S, et al. Soft-sensor development using correlation-based just-in-time modeling[J]. AIChE Journal, 2009, 55(7):1754-1765.
[19]	刘乙奇, 黄道平, 李艳. 基于改进JIT算法的软测量建模及其在污水处理中的应用[J]. 华南理工大学学报(自然科学版), 2011, 39(5):55-60+67. LIU Y Q, HUANG D P, LI Y. Enhanced JIT-based soft-sensing modeling and its application to wastewater treatment[J]. Journal of South China University of Technology(Natural Science Edition), 2011, 39(5):55-60+67.
[20]	KANEKO H, ARAKAWA M, FUNATSU K. Development of a new soft sensor method using independent component analysis and partial least squares[J]. AIChE J., 2009, 55(1):87-98.
[21]	KANEKO H, ARAKAWA M, FUNATSU K. Applicability domains and accuracy of prediction of soft sensor models[J]. AIChE J., 2011, 57(6):1506-1513.
[22]	ANDERSEN C M, BRO R. Variable selection in regression-a tutorial[J]. Journal of Chemometrics, 2010, 24(11/12):728-737.
[23]	NORGAARD L, SAUDLAND A, WAGNER J, et al. Interval partial least-squares regression (iPLS):a comparative chemometric study with an example from near-infrared spectroscopy[J]. Applied Spectroscopy, 2000, 54(3):413-419.
[24]	MEHMOOD T, LILAND K H, SNIPEN L, et al. A review of variable selection methods in Partial Least Squares Regression[J]. Chemometrics & Intelligent Laboratory Systems, 2012, 118(16):62-69.
[25]	KANEKO H, FUNATSU K. Maintenance-free soft sensor models with time difference of process variables[J]. Chemometrics & Intelligent Laboratory Systems, 2011, 107(2):312-317.
[26]	KANEKO H, FUNATSU K. Automatic determination method based on cross-validation for optimal intervals of time difference[J]. Journal of Chemical Engineering of Japan, 2013, 46(3):219-225.
[27]	KANEKO H, FUNATSU K. Discussion on time difference models and intervals of time difference for application of soft sensors[J]. Industrial and Engineering Chemistry Research, 2013, 52(3):1322-1334.
[28]	ERIKSSON L, JOHANSSON E, ANTTI H, et al. Multi-and Megavariate Data Analysis[M]//Metabonomics in Toxicity Assessment. Umeå:Umetrics Acodemy, 2001:362.
[29]	CHONG I G, JUN C H. Performance of some variable selection methods when multicollinearity is present[J]. Chemometrics & Intelligent Laboratory Systems, 2005, 78(1):103-112.
[30]	GOSSELIN R, RODRIGUE D, DUCHESNE C. A Bootstrap-VIP approach for selecting wavelength intervals in spectral imaging applications[J]. Chemometrics & Intelligent Laboratory Systems, 2010, 100(1):12-21.