自确认软测量模型研究及其在污水处理中的应用
|
摘要
随着社会的发展，对现代工业过程建模与控制的要求日益提高，而现代工业生产过程越来越复杂，往往存在着参数时变、多变量耦合、强非线性、大滞后等特点，面对这些特点，传统传感器无法得到有效应用，重要变量无法得到快速精确测量，生化过程无法得到有效优化和诊断。由此，软测量作为现代复杂过程工业中较难甚至无法由硬件在线检测参量实时估计的有效手段，也是现代过程控制领域的研究热点之一，受到了国内外学者的广泛关注。本文以污水处理过程为背景，结合其生化和工艺知识，对软测量的若干建模方法进行了研究，克服了软测量模型在线辨识，在线自诊断，在线自修复，复杂过程建模等难题，并在结合多方面数据重构知识和硬件仪表自确认技术的基础上提出了自确认软测量的概念，同时文中不仅仅关注了复杂污水处理过程的建模，更是首次在软测量预测模型的基础上实现了污水管网投药的控制，最后对软测量在污水处理工业中的实践应用进行了充分的探讨。
本文主要研究如下：
1.针对污水处理过程特点，针对性的应用了两种局部学习模型即RBF (RadicalBasis Function)神经网络和LWPR (Locally Weighted Projection Regression)进行软测量建模，避免了全局模型所带来的训练参数多，模型结构相对复杂等问题，并且在软测量数据预处理中还采用了PCA (Principal Component Analysis)和Jolliffe参数相结合的方式，不仅克服了单纯Jolliffe参数离群点检测存在的问题，而且通过PCA降低高维数据的复杂度和相关性为软测量建模提供了便利。所引进的方法用于污水厂出水指标5天生物需氧量（BOD5）的软测量建模，结果表明两种局部模型算法为出水指标BOD5的实时、精确预测提供了一条有效的途径。
2.提出了JIT-PLS（Just-in-time Partial Least square）和JIT-RVC(Just-in-time RandomlyVarying Coefficient)算法。该算法在传统JIT数据选择算法理念的基础上，提出一种改进的鲁棒最近相关算法（RNC, Robust Near Coefficient），其充分利用了相关性最大原则原理，提高了传统JIT算法的数据选择能力；同时，充分考虑RPLS (Recursive Partial LeastSquare)、RVC的本质鲁棒性特性，提高了传统JIT算法的非线性逼近能力、在线动态处理能力和抗干扰能力；此外，为了避免数据的老化和新数据不断加入对JIT算法数据选择所带来的负担，在该算法中还引入了移动窗口技术。基于实际过程数据仿真结果表明，预测结果与化验室结果吻合，预测精度高，有效的加强了软测量的在线学习能力和污水生化处理的BOD5软测量在线预测能力。
3.提出了新型的软测量模型即自确认软测量模型。在自确认软测量模型中，为了确保输入传感器数据的可靠性，充分利用PCA对输入传感器数据进行在线校验和故障重构，同时，利用ISS (Input Sensor Status)来指示当前输入传感器的状态。另外，通过新的数据选择方法和Ensemble学习对JIT学习算法进行了改进。此外，通过输出方差和ICP (Inductive Cofidence Predictor)两种方法对软测量输出的不确定性进行了描述。最后，传统预测模型的功能是在模型的基础上利用可测变量对控制对象的某些不可测或难于实时测量的变量值或状态值进行预测，本章将研究预测模型的综合输出机制，以更好的反应控制对象的状态，模型的输出将不是单个预测值的输出，而是同时输出四种信息：带概率区间的输出、模型的状态（故障状态，重构状态和迷失状态等等）、不确定性、故障信息和校验输出值，并对发生故障的输入传感器进行数据重构和修复以达到模型自校验和自诊断的效应。本论文的研究工作引入不确定的描述方法，一方面用不确定性来描述预测值的正常与否，另一方面利用不确定性区间还能对被预测的变量的进行校验。其中，特别需要指出的是课题创新性的运用鲁棒统计学中的CP (Cofidence Predictor)技术对预测模型的不确定进行定性的描述。此外，可以将输入传感器的状态也作为模型的输出，从而达到对预测模型那个从输入，模型和输出的三方位检测，使得模型的输出不再只是单纯的预测值，还有不确定区间和输入传感器的输入状态。实验仿真结果表明，SEVA (Self-valicating)软测量的RMSE (Root Mean Square Error)得到了大幅提高，即使在辅助变量传感器发生故障的情况下。此外，SEVA软测量还能用于校验在线仪表，这些在污水厂的出水指标预测中得到了进一步的验证。
4.将软测量模型拓展到了前馈控制方案中，提出了基于ARMA (Auto-Regressiveand Moving Average Model)软测量模型的前馈控制方案，该方法既可以对泵站的输入流量做到长达6小时的预测，同时也可以进一步优化前馈控制方案，同时，针对污水管网系统输入流量的特性和用户用水特点，提出了日用水的均值的管网入水流量数据标准化方法。总体方案在优化管网系统的投药系统得到了有效应用，无论是在仿真平台还是在澳大利亚Bellambi泵站都得到了有效验证，在一定程度上实现了管网的节能减排效果。
最后对论文中所作的研究进行简要总结，并指出了这一领域有待进一步深入研究的问题。
With the ever-increasing development, industry system operation and maintenance arebecoming more complicated with the features of coupled variables, significant nonlinearities,parameters shift and time delay, thereby resulting in more and more rigrid requirements onprocess modeling and control. However, traditional sensors are unavailable and criticalvariables can not detect efficiently in a real-time way, leading to imposibilty of efficientprocess optimization and diagnosis. Soft sensors are a good alternative in a complex processin response to lack of efficient sensors to detect critical parameters. Against the background ofwastewater treatment, this paper focuses on soft sensors modelling together with theknowledge of chemical and biological reactions, further presents self-validating soft sensorsconcept by combining data reconstruction and hardware sensor self-validating techniquewhich addresses the issues of soft sensor model on-line identification, on-line self-validation,on-line self-reconstruction and complex process modelling. In addition, this paper is the firstattempt to control chemical dosage control in sewer networks. Main results have beenobtained as follows:
1. To address the problems exsting in wastewater process and inefficiency of globalmodel, two local model learning methods, i.e., the RBF neural network and the LWPRalgorithm, are presented. PCA and Jolliffe parameters are combining as data pre-treatmentmothod, not only addressing the issues existing in pure Jolliffe parameters but also makingfull use of PCA to descrease the dimentions of input data. Due to significant time delay whenusing tradictional sensors in WWTP, in particular, BOD5dectection requires5-10daysanalysis, the simulation results showed that both local models performe well with a excellentaccuracy, provding more potential to use local modeling methods to soft sensor modelling.
2. To improve ability of JIT learning, JIT-PLS and JIT-RVC algorithms are proposed inthis section. Their data selection method is enhanced by robust nearest algorithm and biascompensation algorithm. By doing so, the abilities of nonlinear tracking, dynamicalprocessing and resisting noises are improved dramatically. The simulation shows that theenhanced JIT algorithems outperform conventional JIT and RPLS, it not only improves theprediction accuracy but also enhances the adaptability and robustness of soft sensors.
3. To overcome the reliable problems of soft sensors, dissimilar to traditional soft sensors,we propose an integrated framework known as SEVA soft sensors. This will perform asfollows: Validate the input sensors before making a prediction. A PCA model proposed byDunia et al. is obtained to detect, identify and reconstruct a faulty input sensor. Meanwhile, this PCA model can be further utilized in the sequential data selection for JIT learning; Builda JIT-ENS prediction model for output variables on the basis of the validated inputs, ratherthan the raw input variables, thus ensuring the prediction model is well conditioned; Validatethe prediction values of the JIT-ENS model by utilizing confidence intervals. Theseconfidence intervals, obtained from ICP algorithm, can characterize the uncertainty of theprediction model and provide further useful information about prediction quality; Generateanother three types of outputs for a soft sensor: Input Sensor Status (ISS), Output sensorstatus (OSS) and Validated Measurement (VM), beyond uncertainty values (UV) andPrediction Values (PV). The usefulness of the proposed SEVA soft sensors is demonstratedthrough a case study of a wastewater treatment process.
4. To optimize chemical dosage in the sewer network, ARMA-based soft sensor isproposed in this section. It is not only capable of making6hours ahead prediction, but also isfurther used to improve feedforward control system. Therefore, the chemical dosage cost issaved dramatically. The methology is implemented in both SewerX model platform andBellambi pump station, further showing the efficiency of ARMA model-based feedforwardcontrol strategy.
Finally, the summaries are obtained and pay the way for further research.
本文主要研究如下：
1.针对污水处理过程特点，针对性的应用了两种局部学习模型即RBF (RadicalBasis Function)神经网络和LWPR (Locally Weighted Projection Regression)进行软测量建模，避免了全局模型所带来的训练参数多，模型结构相对复杂等问题，并且在软测量数据预处理中还采用了PCA (Principal Component Analysis)和Jolliffe参数相结合的方式，不仅克服了单纯Jolliffe参数离群点检测存在的问题，而且通过PCA降低高维数据的复杂度和相关性为软测量建模提供了便利。所引进的方法用于污水厂出水指标5天生物需氧量（BOD5）的软测量建模，结果表明两种局部模型算法为出水指标BOD5的实时、精确预测提供了一条有效的途径。
2.提出了JIT-PLS（Just-in-time Partial Least square）和JIT-RVC(Just-in-time RandomlyVarying Coefficient)算法。该算法在传统JIT数据选择算法理念的基础上，提出一种改进的鲁棒最近相关算法（RNC, Robust Near Coefficient），其充分利用了相关性最大原则原理，提高了传统JIT算法的数据选择能力；同时，充分考虑RPLS (Recursive Partial LeastSquare)、RVC的本质鲁棒性特性，提高了传统JIT算法的非线性逼近能力、在线动态处理能力和抗干扰能力；此外，为了避免数据的老化和新数据不断加入对JIT算法数据选择所带来的负担，在该算法中还引入了移动窗口技术。基于实际过程数据仿真结果表明，预测结果与化验室结果吻合，预测精度高，有效的加强了软测量的在线学习能力和污水生化处理的BOD5软测量在线预测能力。
3.提出了新型的软测量模型即自确认软测量模型。在自确认软测量模型中，为了确保输入传感器数据的可靠性，充分利用PCA对输入传感器数据进行在线校验和故障重构，同时，利用ISS (Input Sensor Status)来指示当前输入传感器的状态。另外，通过新的数据选择方法和Ensemble学习对JIT学习算法进行了改进。此外，通过输出方差和ICP (Inductive Cofidence Predictor)两种方法对软测量输出的不确定性进行了描述。最后，传统预测模型的功能是在模型的基础上利用可测变量对控制对象的某些不可测或难于实时测量的变量值或状态值进行预测，本章将研究预测模型的综合输出机制，以更好的反应控制对象的状态，模型的输出将不是单个预测值的输出，而是同时输出四种信息：带概率区间的输出、模型的状态（故障状态，重构状态和迷失状态等等）、不确定性、故障信息和校验输出值，并对发生故障的输入传感器进行数据重构和修复以达到模型自校验和自诊断的效应。本论文的研究工作引入不确定的描述方法，一方面用不确定性来描述预测值的正常与否，另一方面利用不确定性区间还能对被预测的变量的进行校验。其中，特别需要指出的是课题创新性的运用鲁棒统计学中的CP (Cofidence Predictor)技术对预测模型的不确定进行定性的描述。此外，可以将输入传感器的状态也作为模型的输出，从而达到对预测模型那个从输入，模型和输出的三方位检测，使得模型的输出不再只是单纯的预测值，还有不确定区间和输入传感器的输入状态。实验仿真结果表明，SEVA (Self-valicating)软测量的RMSE (Root Mean Square Error)得到了大幅提高，即使在辅助变量传感器发生故障的情况下。此外，SEVA软测量还能用于校验在线仪表，这些在污水厂的出水指标预测中得到了进一步的验证。
4.将软测量模型拓展到了前馈控制方案中，提出了基于ARMA (Auto-Regressiveand Moving Average Model)软测量模型的前馈控制方案，该方法既可以对泵站的输入流量做到长达6小时的预测，同时也可以进一步优化前馈控制方案，同时，针对污水管网系统输入流量的特性和用户用水特点，提出了日用水的均值的管网入水流量数据标准化方法。总体方案在优化管网系统的投药系统得到了有效应用，无论是在仿真平台还是在澳大利亚Bellambi泵站都得到了有效验证，在一定程度上实现了管网的节能减排效果。
最后对论文中所作的研究进行简要总结，并指出了这一领域有待进一步深入研究的问题。
With the ever-increasing development, industry system operation and maintenance arebecoming more complicated with the features of coupled variables, significant nonlinearities,parameters shift and time delay, thereby resulting in more and more rigrid requirements onprocess modeling and control. However, traditional sensors are unavailable and criticalvariables can not detect efficiently in a real-time way, leading to imposibilty of efficientprocess optimization and diagnosis. Soft sensors are a good alternative in a complex processin response to lack of efficient sensors to detect critical parameters. Against the background ofwastewater treatment, this paper focuses on soft sensors modelling together with theknowledge of chemical and biological reactions, further presents self-validating soft sensorsconcept by combining data reconstruction and hardware sensor self-validating techniquewhich addresses the issues of soft sensor model on-line identification, on-line self-validation,on-line self-reconstruction and complex process modelling. In addition, this paper is the firstattempt to control chemical dosage control in sewer networks. Main results have beenobtained as follows:
1. To address the problems exsting in wastewater process and inefficiency of globalmodel, two local model learning methods, i.e., the RBF neural network and the LWPRalgorithm, are presented. PCA and Jolliffe parameters are combining as data pre-treatmentmothod, not only addressing the issues existing in pure Jolliffe parameters but also makingfull use of PCA to descrease the dimentions of input data. Due to significant time delay whenusing tradictional sensors in WWTP, in particular, BOD5dectection requires5-10daysanalysis, the simulation results showed that both local models performe well with a excellentaccuracy, provding more potential to use local modeling methods to soft sensor modelling.
2. To improve ability of JIT learning, JIT-PLS and JIT-RVC algorithms are proposed inthis section. Their data selection method is enhanced by robust nearest algorithm and biascompensation algorithm. By doing so, the abilities of nonlinear tracking, dynamicalprocessing and resisting noises are improved dramatically. The simulation shows that theenhanced JIT algorithems outperform conventional JIT and RPLS, it not only improves theprediction accuracy but also enhances the adaptability and robustness of soft sensors.
3. To overcome the reliable problems of soft sensors, dissimilar to traditional soft sensors,we propose an integrated framework known as SEVA soft sensors. This will perform asfollows: Validate the input sensors before making a prediction. A PCA model proposed byDunia et al. is obtained to detect, identify and reconstruct a faulty input sensor. Meanwhile, this PCA model can be further utilized in the sequential data selection for JIT learning; Builda JIT-ENS prediction model for output variables on the basis of the validated inputs, ratherthan the raw input variables, thus ensuring the prediction model is well conditioned; Validatethe prediction values of the JIT-ENS model by utilizing confidence intervals. Theseconfidence intervals, obtained from ICP algorithm, can characterize the uncertainty of theprediction model and provide further useful information about prediction quality; Generateanother three types of outputs for a soft sensor: Input Sensor Status (ISS), Output sensorstatus (OSS) and Validated Measurement (VM), beyond uncertainty values (UV) andPrediction Values (PV). The usefulness of the proposed SEVA soft sensors is demonstratedthrough a case study of a wastewater treatment process.
4. To optimize chemical dosage in the sewer network, ARMA-based soft sensor isproposed in this section. It is not only capable of making6hours ahead prediction, but also isfurther used to improve feedforward control system. Therefore, the chemical dosage cost issaved dramatically. The methology is implemented in both SewerX model platform andBellambi pump station, further showing the efficiency of ARMA model-based feedforwardcontrol strategy.
Finally, the summaries are obtained and pay the way for further research.
引文
[1] Aguado D., et al., Comparison of different predictive models for nutrient estimation in asequencing batch reactor for wastewater treatment [J]. Chemometrics and IntelligentLaboratory Systems,2006,84(1-2):75-81
[2] Jeppson U., et al., Benchmark Simulation no.2: general protocol and exploratory casestudies [J]. Water Sci. Technol.,2007,56(8):67-68
[3] Huang M. Z., et al., A fast predicting neural fuzzy model for on-line estimation of nutrientdynamics in an anoxic/oxic process [J]. Bioresource Technology,2010,101(6):1642-1651
[4] Yan W., Shao H. and Wang X., Soft sensing modeling based on support vector machineand Bayesian model selection [J]. Computers&Chemical Engineering,2004,28(8):1489-1498
[5] Ahmed F., Nazir S. and Yeo Y. K., A recursive PLS-based soft sensor for prediction of themelt index during grade change operations in HDPE plant [J]. Korean Journal ofChemical Engineering,2009,26(1):14-20
[6] Facco P., Bezzo F. and Barolo M., Nearest-Neighbor Method for the AutomaticMaintenance of Multivariate Statistical Soft Sensors in Batch Processing [J]. Industrial&Engineering Chemistry Research,2010,49(5):2336-2347
[7] Martens H., et al., Analysis of designed experiments by stabilised PLS Regression andjack-knifing [J]. Chemometrics and Intelligent Laboratory Systems,2001.58(2):151-170
[8] Joe Qin, S., Recursive PLS algorithms for adaptive data modeling. Computers&Chemical Engineering,1998.22(4-5): p.503-514.
[9] Liu X., et al., Moving window kernel PCA for adaptive monitoring of nonlinear processes[J]. Chemometrics and Intelligent Laboratory Systems,2009,96(2):132-143
[10] Wang X., Kruger U. and Irwin G. W., Process Monitoring Approach Using Fast MovingWindow PCA [J]. Industrial&Engineering Chemistry Research,2005,44(15):5691-5702
[11] He X. B. and Yang Y. P., Variable MWPCA for Adaptive Process Monitoring [J].Industrial&Engineering Chemistry Research,2007.47(2):419-427
[12] Zhao C., Wang F. and Jia M., Dissimilarity analysis based batch process monitoringusing moving windows [J]. Aiche Journal,2007,53(5):1267-1277
[13] Dayal B. S. and MacGregor J. F., improved PLS algorithms [J]. Journal of Chemometrics,1997,11(1):73-85
[14] Li W., et al., Recursive PCA for adaptive process monitoring [J]. Journal of ProcessControl,2000,10(5):471-486
[15] Fortescue T. R., Kershenbaum L. S. and Ydstie B. E., Implementation of self-tuningregulators with variable forgetting factors [J]. Automatica,1981,17(6):831-835.
[16] Choi S. W., et al., Adaptive Multivariate Statistical Process Control for MonitoringTime-Varying Processes [J]. Industrial&Engineering Chemistry Research,2006,45(9):3108-3118
[17] Krogh A. and Sollich P., Statistical mechanics of ensemble learning [J]. Physical ReviewE,1997,55(1):811-815
[18] Krogh A. and Vedelsby J., Neural network ensembles cross validation and active learning[J]. Advanced in Neural Information Processing Systems,1995.5(7):231-238.
[19] Liu J., On-line soft sensor for polyethylene process with multiple production grades [J].Control Engineering Practice,2007,15(7):769-778
[20] Kadlec P., On robust and adaptive soft sensors [D]. London: School of DesignEngineering&Computing. Bournemouth University,2009
[21] Kadlec P. and Gabrys B., Adaptive Local Learning Soft Sensor for Inferential ControlSupport [A]. International Conference on Computational Intelligence for ModellingControl&Automation,2008
[22] Huang Z. Y. and Mei C. L., Soft sensor modeling using SVR based on Genetic Algorithmand Akaike Information Criterion [A]. International Conference on IntelligentHuman-Machine Systems and Cybernetics,2009,2:123-127.
[23] Clarke D. W. and Fraher P. M. A., Model-based validation of a DOx sensor [J]. ControlEngineering Practice,1996,4(9):1313-1320
[24] Henry, M. P. and Clarke D. W., The self-validating sensor: rationale, definitions andexamples [J]. Control Engineering Practice,1993,1(4):585-610
[25]冯志刚,自确认压力传感器的研究[D].哈尔滨:哈尔滨工业大学自动化测试与控制系,2009
[26]高羽,自确认传感器理论及应用研究[D].上海:复旦大学信息科学与工程学院,2008
[27] Qin S. J., Yue H. and Dunia R., Self-Validating Inferential Sensors with Application toAir Emission Monitoring [J]. Industrial&Engineering Chemistry Research,1997,36(5):1675-1685
[28] Wang S. and Chen Y., Sensor validation and reconstruction for building central chillingsystems based on principal component analysis [J]. Energy Conversion and Management,2004,45(5):673-695.
[29] Liu J., Chen D. S. and Shen J. F., Development of Self-Validating Soft Sensors UsingFast Moving Window Partial Least Squares [J]. Industrial&Engineering ChemistryResearch,2010,49(2):11530-11546
[30] Olsson G, Nielsen M and Yuan. Z. G. e. al, Instrumentation, Control and Automation inWastewater Systems [M]. London: Hove, IWA,2005:33-38
[31] Brosilow C. and Tong M., Inferential control of processes: Part II. The structure anddynamics of inferential control systems [J]. Aiche Journal,1978,24(3):492-500
[32] Xuefeng Z., Soft-sensing Technique and Its Application [J]. Journal of South ChinaUniversity of Science Technology,2002,30(11):61-67
[33]Fortuna L., Soft Sensors for Monitoring and Control of Industrial Processes [M]. London:Springer Verlag,2007:37-70
[34] Wan J., et al, An application of artificial neuromolecular system for effluent qualityprediction of wastewater treatment plant [J]. Chinese Instrument EnvironmentEngineering,2000,10(3):155-162
[35] Rodríguez J., Ruiz G., Molina F., Roca E. and Lema J. M., A hydrogen-basedvariable-gain controller for anaerobic digestion processes [J]. Water Science&Technology,2006,54(2):57-62
[36] Sanchez E. N., Gonzalez J. M., and Ramirez E.. Minimal PD fuzzy control of awastewater treatment plant [A]. Proceedings of the2000IEEE International Symposiumon Intelligent Control [C]. California: IEEE,2000:200-215
[37] Bisschops I, et al., Automatic detection of exogenous respiration end-point usingartificial neural network [J]. Water Science and Technology,2006,53(4-5):273-281
[38] Pires O. C., et al., Knowledge-based fuzzy system for diagnosis and control of anintegrated biological wastewater treatment process [J]. Water Science&Technology,2006,53(4-5):313-320
[39] Weili, et al., Soft-measuring Technique to Predict BOD Based on PCA-GABP NeuralNetworks [J]. Control Engineering of China,2004,19(12):78-82
[40] Aguado D., et al., Using SOM and PCA for analysing and interpreting data from aP-removal SBR [J]. Engineering Applications of Artificial Intelligence,2008,21(6):919-930
[41] Lee M. W., et al., Real-time remote monitoring of small-scaled biological wastewatertreatment plants by a multivariate statistical process control and neural network-basedsoftware sensors [J]. Process Biochemistry,2008,43(10):1107-1113
[42] Choi D. J. and Park H., A hybrid artificial neural network as a software sensor foroptimal control of a wastewater treatment process [J]. Water Research,2001,35(16):3959-3967
[43] Cao L. J., et al., A comparison of PCA, KPCA and ICA for dimensionality reduction insupport vector machine [J]. Neurocomputing,2003,55(1-2):321-336.
[44] Lee D. S., et al., Multivariate Online Monitoring of a Full-Scale Biological AnaerobicFilter Process Using Kernel-Based Algorithms [J]. Industrial&Engineering ChemistryResearch,2006,45(12):4335-4344
[45] Choi S. W. and Lee I. B., Multiblock PLS-based localized process diagnosis [J]. Journalof Process Control,2005,15(3):295-306
[46] Chiang L. H., Pell R. J., and Seasholtz M. B., Exploring process data with the use ofrobust outlier detection algorithms [J]. Journal of Process Control,2003,13(5):437-449
[47] Jolliffe I. T., Principal Component Analysis,3nd [M]. New York: Springer,2002:28-80
[48] Henze M., et al., Activated sludge model No.1[R]. London: IAWQ,1987
[49] Henze M., et al., Activated sludge model No.2[R]. London: IAWQ,1995
[50] Gujor W., et al., Activated sludge model No.3[J]. Water Science and Technology,1999,39(1):183-193
[51] Corriou J. P., et al., Model predictive control of wastewater treatment plants applicationto the BSM1benchmark [J]. Computer Aided Chem. Eng.,2004,18(14):625-634
[52] Mianjie Z., Study on Wastewater Treatment Process Control Strategy Based onBenchmark [D]. Zhejiang: Department of Contorl Science and Engineering, ZhejiangUniversity,2006
[53] Wenhao L. D. S., Set Up BSM1with Matlab/Simulink. Environments Science&Technology,2007.30(11): p.86-89.
[54] Vrecko D., et al, Feedforward-feedback control of an activated sludge process: asimulation study [J]. Water Science and Technology,2003,47(12):19-26
[55] Benedetti L., et al., Multi-criteria analysis of wastewater treatment plant design andcontrol scenarios under uncertainty [J]. Environmental Modelling&Software,2010,25(5):616-621
[56] Yongzai L. et al., Soft Sensor for Total Nitrogen in Waster Water Based on RBF-NNOptimized by Hierarchical GA [J]. Computer Simulation,2008.25(7):137-140
[57] Spérandio M., and Espinosa M. C., Modelling an aerobic submerged membranebioreactor with ASM models on a large range of sludge retention time [J]. Desalination,2008,231(1-3):82-90
[58] Sotomayor O. A. Z., Park S. W., and Garcia C., Software sensor for on-line estimation ofthe microbial activity in activated sludge systems [J]. ISA Transactions,2002:127-143
[59] Theilliol, D., et al., On-line estimation of unmeasured inputs for anaerobic wastewatertreatment processes [J]. Control Engineering Practice,2003,11(9):1007-1019
[60] Bogaerts P., and Wouwer A. V., Software sensors for bioprocesses [J]. ISA Transactions,2003,42(4):547-558
[61] Alcaraz-González V., et al., Software sensors for highly uncertain WWTPs: a newapproach based on interval observers [J]. Water Research,2002,36(10):2515-2524
[62] Aguado D., et al., A methodology for sequencing batch reactor identification withartificial neural networks: A case study [J]. Computers&Chemical Engineering,2009,33(2):465-472
[63] Zyngier D., et al., Robust soft sensors for SBR monitoring [J]. Water Science andTechnology,2001,43(3):101-105
[64] Feitkenhauer H. and Meyer U., Software sensors based on titrimetric techniques for themonitoring and control of aerobic and anaerobic bioreactors [J]. BiochemicalEngineering Journal,2004,17(2):147-151
[65] Zyngier D., et al., Soft sensors with white-and black-box approaches for a wastewatertreatment process [J]. Brazilian Journal of Chemical Engineering,2000,17:433-440
[66] Zyngier D., et al., Hybrid estimator for a wastewater treatment process [A]. The2ndCongress on Process Systems Engineering for the MERCOSUR FLORIANOPOLIS [C].BRAZIL: LATIN AMERICAN APPLIED RESEARCH,2001:507-511
[67] Onkal Engin, et al., Determination of the relationship between sewage odour and BODby neural networks [J]. Environmental Modelling&Software,2005,20(7):843-850
[68] Qian L., Neural Network Optimization Based on Hierarchical Genetic Algorithm and Itsapplication in Building Soft-sensor for Nitrogen Removal Process in WastewaterTreatment [D]. Shanghai: Department of Control Scicence and Engineering, ShanghaiJiao Tong University,2008
[69] Li Minghe and ming Y., Research and Application of Wastewater TN Soft-sensor Model[J]. Automation&Instrumentation,2009,9:1-4
[70] Pai T. Y., et al., Using fuzzy inference system to improve neural network for predictinghospital wastewater treatment plant effluent [J]. Computers&Chemical Engineering,2009,33(7):1272-1278
[71] Perendeci A., et al., Prediction of effluent quality of an anaerobic treatment plant underunsteady state through ANFIS modeling with on-line input variables [J]. ChemicalEngineering Journal,2008,145(1):78-85
[72] Perendeci A., et al., Effects of phase vector and history extension on prediction power ofadaptive-network based fuzzy inference system (ANFIS) model for a real scale anaerobicwastewater treatment plant operating under unsteady state [J]. Bioresource Technology,2009,100(20):4579-4587
[73] Fortuna L., Graziani S. and Xibilia M. G., Soft sensors for product quality monitoring indebutanizer distillation columns [J]. Control Engineering Practice,2005,13(4):499-508
[74]Li S. J., et al., Soft sensing modeling via artificial neural network based PSO-Alopex [A].Proceedings of2005international conference on machine learning and cybernetics [C].Guangzhou: IEEE,2005:38-43
[75] Alhoniemi E. S. A., Process monitoring and modeling using the self-organizing map [J].Integrated Computer aided Engineering,1999,6(1):3-14
[76] Avoy T. M., Intelligent "control" applications in the process industries [J]. AnnualReviews in Control,2002,26(1):75-86
[77] Ferrer A., et al., PLS: A versatile tool for industrial process improvement andoptimization [J]. Applied Stochastic Models in Business and Industry,2008,24(6):551-567
[78] Rosen C., A Chemometric Approach to Process Monitoring and Control withApplications to Wastewater Treatment Operation [D]. Sweden: Lund Institute ofTechnology,2001
[79] Le Bonté, et al., Toxic event detection by respirometry and adaptive principalcomponents analysis [J]. Environmetrics,2005,16(6):589-601
[80] Lee J. M., Yoo C. and Lee I. B., On-line batch process monitoring using a consecutivelyupdated multiway principal component analysis model [J]. Computers&ChemicalEngineering,2003,27(12):1903-1912
[81] Villez K., Multivariate and qualitative data analysis for monitoring, diagnosis and controlof sequencing batch reactors for wastewater treatment [D]. Belgium: University of Gent,2007
[82] Yoo C. K., Lee D. S. and Vanrolleghem P. A., Application of mutiway ICA for onlineprocess monitoring of a sequencing batch reactor [J]. Water Research,2004,38(7):1715-1732
[83] Durante C., et al., Application of N-PLS to gas chromatographic and sensory data oftraditional balsamic vinegars of modena [J]. Chemometrics and Intelligent LaboratorySystems,2006,83:54-65
[84] Serra P., et al., Development of a real-time expert system for wastewater treatment plantscontrol [J]. Control Engineering Practice,1993,1(2):329-335
[85] Yu R. F., et al., Dynamic control of disinfection for wastewater reuse applying ORP/pHmonitoring and artificial neural networks [J]. Resources, Conservation and Recycling,2008,52(8-9):1015-1021
[86] El-Din A. G. and Smith D. W., A neural network model to predict the wastewater inflowincorporating rainfall events [J]. Water Research,2002,36(5):1115-1126
[87] Berthouex P. M. and Polkowski L. B., Optimum Waste Treatment Plant Design underUncertainty [J]. Water Pollution Control Federation,1970,42(9):1589-1613
[88] Qin X. S., et al, The application of genetic algorithms on grey nonlinear programming inwater environment [J]. Advances in Water Science,2002,13(1):31-36
[89] Xianmin, S. and Weisheng Y., SVI Soft Sensor Design and Modelling [J]. Industry andMine Automation,2007,27(3):27-29
[90] Lumley D., On-line instrument confirmation: how can we check that our instruments areworking?[J]. Water Science and Technology,2002,45(4-5):469-476
[91] Woo S. H., et al., On-line estimation of key process variables based on kernel partialleast squares in an industrial cokes wastewater treatment plant [J]. Journal of HazardousMaterials,2009,161(1):538-544
[92] Zyngier D., et al., Soft sensor development and experimental application to a wastewatertreatment process [J]. Computer Aided Chemical Engineering,8:943-948
[93] Dianlin D., et al. Delay Time Identification and Dynamic Characteristics Study on ANNSoft Sensor [A]. The Sixth International Conference on Intelligent Systems Design andApplications [C]. Jinan:2006:637-642
[94] Warne K., et al., Statistical and computational intelligence techniques for inferentialmodel development: a comparative evaluation and a novel proposition for fusion [J].Engineering Applications of Artificial Intelligence,2004,17(8):871-885
[95] Vijayakumar S. and Schaal S., Incremental Online Learning in High Dimensions [J].Neural Computation,2005,17:2602-2634
[96] Huang M. Z., et al., A fast predicting neural fuzzy model for on-line estimation ofnutrient dynamics in an anoxic/oxic process [J]. Bioresource Technology,2010,101(6):1642-1651
[97] Edakunni N. U., Bayesian Locally Weighted Online Learning [D]. Edinburgh: School ofInformatics, University of Edinburgh,2009
[98] Fujiwara K., Kano M. and Hasebe S., Development of correlation-based patternrecognition and its application to adaptive soft-sensor design [A].2009ICCAS-SICE [C].Fukuoka: IEEE,2009
[99] Fujiwara K., et al., Soft-sensor development using correlation-based just-in-timemodeling [J]. Aiche Journal,2009,55(7):1754-1765
[100] Cheng C. and Chiu M. S., A new data-based methodology for nonlinear processmodeling [J]. Chemical Engineering Science,2004,59(13):2801-2810
[101] Edakunni N. U., Schaal S., and Vijayakumar S., Kernel Carpentry for OnlineRegression using Randomly Varying Coefficient Model [A]. Proceedings of theInternational Joint Conference on Artificial Intelligence [C]. Hyderabad2007
[102] Guang-Bin H., Lei C. and Chee-Kheong S., Universal approximation using incrementalconstructive feedforward networks with random hidden nodes [J]. IEEE Transactionson Neural Networks,200617(4):879-892
[103] Khayamian T., Robustness of PARAFAC and N-PLS regression models in relation tohomoscedastic and heteroscedastic noise [J]. Chemometrics and Intelligent LaboratorySystems,2007,88:35-40
[104] Kadlec P., Gabrys B. and Strandt S., Data-driven Soft Sensors in the process industry[J]. Computers and Chemical Engineering,2009,33(1):795-814
[105] Yue H. H. and Qin S. J., Reconstruction-Based Fault Identification Using a CombinedIndex [J]. Industrial&Engineering Chemistry Research,2001,40(20):4403-4414
[106] Tien D. X., Lim K. W. and Jun L., Comparative study of PCA approaches in processmonitoring and fault detection [A].30th Annual Conference of IEEE in IndustrialElectronics Society [C]. Busan: IEEE,2004
[107] Zhang F. A mixture probabilistic PCA model for multivariate processes monitoring [A].American Control Conference [C]. Washington D. C.: IEEE,2004
[108] Pessel N., et al. An improved PCA fault detection for the diagnosis [A].9th WSEASInternational Conference on Automatic Control, Modeling and Simulation [C]. Istanbul:World Scientific and Engineering Academy and Society,2007
[109] Martin E. B. and Morris A. J., Non-parametric confidence bounds for processperformance monitoring charts [J]. Journal of Process Control,1996,6(6):349-358
[110] Ciba J., et al., PLS-EP algorithm to predict aluminum content in soils of BeskidMountains region [J]. Chemosphere,2009,76(4):565-571
[111] Facco P., et al., Moving average PLS soft sensor for online product quality estimation inan industrial batch polymerization process [J]. Journal of Process Control,2009,19(3):520-529
[112] Cheng C. and Chiu M. S., Nonlinear process monitoring using JITL-PCA [J].Chemometrics and Intelligent Laboratory Systems,2005,76(1):1-13
[113] Merkwirth C., Wichard J. and Ogorzalek M.. A software toolbox for constructingensembles of heterogenous linear and nonlinear models [A]. Proceedings of the2005European Conference on Circuit Theory and Design [C]. Carlo: IEEE,2005
[114] Wichard J. D., Model Selection in an Ensemble Framework [A]. International JointConference on Neural Networks [C]. Vancouver: Computational Intelligent Society,2006
[115] Wichard J. D. and Ogorzalek M.. Time series prediction with ensemble models [A].International Joint Conference on Neural Networks [C]. Vancouver: ComputationalIntelligent Society,2006
[116] Dunia R., et al., Identification of faulty sensors using principal component analysis [J].Aiche Journal,1996,42(10):2797-2812
[117] Papadopoulos H., et al., Inductive Confidence Machines for Regression [A]. ECML2002in Machine Learning [C].2002, Berlin: Springer,185-194
[118] Rallo R. and Giralt F., Best Feature Selection and Data Completion for the Design ofSoft Neural Sensors [A]. Proceedings of AIChE2003,2ndTopical Conference onSensors [C]. San Francisco: ACS,2003
[119] Gernaey K. V., et al., Dynamic influent pollutant disturbance scenario generation usinga phenomenological modelling approach [J]. Environmental Modelling&Software,2011,26(11):1255-1267
[120] Hernebring C, et al., Dynamic online sewer modelling in Helsingborg [J]. WaterScience&Technology,2002,45(4-5):429-436
[121] Langergraber G., et al., Generation of diurnal variation for influent data for dynamicsimulation [J]. Water Science&Technology,2008,57(9):1483-1486
[122] Brockwell P. J. and Davis R. A., Time series: theory and methods [M]. Berlin: SpringerVerlag,2009
[123] Ljung L., System identification [M]. New York: Wiley Online Library,1999
[124] Stoica P. and Nehorai A., On multistep prediction error methods for time series models[J]. Journal of Forecasting,1989,8(4):357-368
[125] Box G. E. P., et al.,, Time series analysis [M]. San Francisco: Holden-day,1970
[126] Isermann R., Parameter adaptive control algorithms-A tutorial [J]. Automatica,1982,18(5):513-528
[127] Sharma K., et al., Predicting hydrogen sulfide formation in sewers: a new model [J].Water,2008,35(2):132-137
[128] Gutierrez O., et al., Effects of long-term pH elevation on the sulfate-reducing andmethanogenic activities of anaerobic sewer biofilms [J]. Water Research,2009,43(9):2549-2557
[129] Ganigué R., et al. Magnesium hydroxide on-line dosing control for sulfide mitigation insewers [A]. Australia's National Water Conference and Exhibtion [C]. Sydney: AWA,2012
[130] Shumway R. H. and Stoffer D. S., Time series analysis and its application [M].3rdEdition ed. Berlin: Springer,2010
[2] Jeppson U., et al., Benchmark Simulation no.2: general protocol and exploratory casestudies [J]. Water Sci. Technol.,2007,56(8):67-68
[3] Huang M. Z., et al., A fast predicting neural fuzzy model for on-line estimation of nutrientdynamics in an anoxic/oxic process [J]. Bioresource Technology,2010,101(6):1642-1651
[4] Yan W., Shao H. and Wang X., Soft sensing modeling based on support vector machineand Bayesian model selection [J]. Computers&Chemical Engineering,2004,28(8):1489-1498
[5] Ahmed F., Nazir S. and Yeo Y. K., A recursive PLS-based soft sensor for prediction of themelt index during grade change operations in HDPE plant [J]. Korean Journal ofChemical Engineering,2009,26(1):14-20
[6] Facco P., Bezzo F. and Barolo M., Nearest-Neighbor Method for the AutomaticMaintenance of Multivariate Statistical Soft Sensors in Batch Processing [J]. Industrial&Engineering Chemistry Research,2010,49(5):2336-2347
[7] Martens H., et al., Analysis of designed experiments by stabilised PLS Regression andjack-knifing [J]. Chemometrics and Intelligent Laboratory Systems,2001.58(2):151-170
[8] Joe Qin, S., Recursive PLS algorithms for adaptive data modeling. Computers&Chemical Engineering,1998.22(4-5): p.503-514.
[9] Liu X., et al., Moving window kernel PCA for adaptive monitoring of nonlinear processes[J]. Chemometrics and Intelligent Laboratory Systems,2009,96(2):132-143
[10] Wang X., Kruger U. and Irwin G. W., Process Monitoring Approach Using Fast MovingWindow PCA [J]. Industrial&Engineering Chemistry Research,2005,44(15):5691-5702
[11] He X. B. and Yang Y. P., Variable MWPCA for Adaptive Process Monitoring [J].Industrial&Engineering Chemistry Research,2007.47(2):419-427
[12] Zhao C., Wang F. and Jia M., Dissimilarity analysis based batch process monitoringusing moving windows [J]. Aiche Journal,2007,53(5):1267-1277
[13] Dayal B. S. and MacGregor J. F., improved PLS algorithms [J]. Journal of Chemometrics,1997,11(1):73-85
[14] Li W., et al., Recursive PCA for adaptive process monitoring [J]. Journal of ProcessControl,2000,10(5):471-486
[15] Fortescue T. R., Kershenbaum L. S. and Ydstie B. E., Implementation of self-tuningregulators with variable forgetting factors [J]. Automatica,1981,17(6):831-835.
[16] Choi S. W., et al., Adaptive Multivariate Statistical Process Control for MonitoringTime-Varying Processes [J]. Industrial&Engineering Chemistry Research,2006,45(9):3108-3118
[17] Krogh A. and Sollich P., Statistical mechanics of ensemble learning [J]. Physical ReviewE,1997,55(1):811-815
[18] Krogh A. and Vedelsby J., Neural network ensembles cross validation and active learning[J]. Advanced in Neural Information Processing Systems,1995.5(7):231-238.
[19] Liu J., On-line soft sensor for polyethylene process with multiple production grades [J].Control Engineering Practice,2007,15(7):769-778
[20] Kadlec P., On robust and adaptive soft sensors [D]. London: School of DesignEngineering&Computing. Bournemouth University,2009
[21] Kadlec P. and Gabrys B., Adaptive Local Learning Soft Sensor for Inferential ControlSupport [A]. International Conference on Computational Intelligence for ModellingControl&Automation,2008
[22] Huang Z. Y. and Mei C. L., Soft sensor modeling using SVR based on Genetic Algorithmand Akaike Information Criterion [A]. International Conference on IntelligentHuman-Machine Systems and Cybernetics,2009,2:123-127.
[23] Clarke D. W. and Fraher P. M. A., Model-based validation of a DOx sensor [J]. ControlEngineering Practice,1996,4(9):1313-1320
[24] Henry, M. P. and Clarke D. W., The self-validating sensor: rationale, definitions andexamples [J]. Control Engineering Practice,1993,1(4):585-610
[25]冯志刚,自确认压力传感器的研究[D].哈尔滨:哈尔滨工业大学自动化测试与控制系,2009
[26]高羽,自确认传感器理论及应用研究[D].上海:复旦大学信息科学与工程学院,2008
[27] Qin S. J., Yue H. and Dunia R., Self-Validating Inferential Sensors with Application toAir Emission Monitoring [J]. Industrial&Engineering Chemistry Research,1997,36(5):1675-1685
[28] Wang S. and Chen Y., Sensor validation and reconstruction for building central chillingsystems based on principal component analysis [J]. Energy Conversion and Management,2004,45(5):673-695.
[29] Liu J., Chen D. S. and Shen J. F., Development of Self-Validating Soft Sensors UsingFast Moving Window Partial Least Squares [J]. Industrial&Engineering ChemistryResearch,2010,49(2):11530-11546
[30] Olsson G, Nielsen M and Yuan. Z. G. e. al, Instrumentation, Control and Automation inWastewater Systems [M]. London: Hove, IWA,2005:33-38
[31] Brosilow C. and Tong M., Inferential control of processes: Part II. The structure anddynamics of inferential control systems [J]. Aiche Journal,1978,24(3):492-500
[32] Xuefeng Z., Soft-sensing Technique and Its Application [J]. Journal of South ChinaUniversity of Science Technology,2002,30(11):61-67
[33]Fortuna L., Soft Sensors for Monitoring and Control of Industrial Processes [M]. London:Springer Verlag,2007:37-70
[34] Wan J., et al, An application of artificial neuromolecular system for effluent qualityprediction of wastewater treatment plant [J]. Chinese Instrument EnvironmentEngineering,2000,10(3):155-162
[35] Rodríguez J., Ruiz G., Molina F., Roca E. and Lema J. M., A hydrogen-basedvariable-gain controller for anaerobic digestion processes [J]. Water Science&Technology,2006,54(2):57-62
[36] Sanchez E. N., Gonzalez J. M., and Ramirez E.. Minimal PD fuzzy control of awastewater treatment plant [A]. Proceedings of the2000IEEE International Symposiumon Intelligent Control [C]. California: IEEE,2000:200-215
[37] Bisschops I, et al., Automatic detection of exogenous respiration end-point usingartificial neural network [J]. Water Science and Technology,2006,53(4-5):273-281
[38] Pires O. C., et al., Knowledge-based fuzzy system for diagnosis and control of anintegrated biological wastewater treatment process [J]. Water Science&Technology,2006,53(4-5):313-320
[39] Weili, et al., Soft-measuring Technique to Predict BOD Based on PCA-GABP NeuralNetworks [J]. Control Engineering of China,2004,19(12):78-82
[40] Aguado D., et al., Using SOM and PCA for analysing and interpreting data from aP-removal SBR [J]. Engineering Applications of Artificial Intelligence,2008,21(6):919-930
[41] Lee M. W., et al., Real-time remote monitoring of small-scaled biological wastewatertreatment plants by a multivariate statistical process control and neural network-basedsoftware sensors [J]. Process Biochemistry,2008,43(10):1107-1113
[42] Choi D. J. and Park H., A hybrid artificial neural network as a software sensor foroptimal control of a wastewater treatment process [J]. Water Research,2001,35(16):3959-3967
[43] Cao L. J., et al., A comparison of PCA, KPCA and ICA for dimensionality reduction insupport vector machine [J]. Neurocomputing,2003,55(1-2):321-336.
[44] Lee D. S., et al., Multivariate Online Monitoring of a Full-Scale Biological AnaerobicFilter Process Using Kernel-Based Algorithms [J]. Industrial&Engineering ChemistryResearch,2006,45(12):4335-4344
[45] Choi S. W. and Lee I. B., Multiblock PLS-based localized process diagnosis [J]. Journalof Process Control,2005,15(3):295-306
[46] Chiang L. H., Pell R. J., and Seasholtz M. B., Exploring process data with the use ofrobust outlier detection algorithms [J]. Journal of Process Control,2003,13(5):437-449
[47] Jolliffe I. T., Principal Component Analysis,3nd [M]. New York: Springer,2002:28-80
[48] Henze M., et al., Activated sludge model No.1[R]. London: IAWQ,1987
[49] Henze M., et al., Activated sludge model No.2[R]. London: IAWQ,1995
[50] Gujor W., et al., Activated sludge model No.3[J]. Water Science and Technology,1999,39(1):183-193
[51] Corriou J. P., et al., Model predictive control of wastewater treatment plants applicationto the BSM1benchmark [J]. Computer Aided Chem. Eng.,2004,18(14):625-634
[52] Mianjie Z., Study on Wastewater Treatment Process Control Strategy Based onBenchmark [D]. Zhejiang: Department of Contorl Science and Engineering, ZhejiangUniversity,2006
[53] Wenhao L. D. S., Set Up BSM1with Matlab/Simulink. Environments Science&Technology,2007.30(11): p.86-89.
[54] Vrecko D., et al, Feedforward-feedback control of an activated sludge process: asimulation study [J]. Water Science and Technology,2003,47(12):19-26
[55] Benedetti L., et al., Multi-criteria analysis of wastewater treatment plant design andcontrol scenarios under uncertainty [J]. Environmental Modelling&Software,2010,25(5):616-621
[56] Yongzai L. et al., Soft Sensor for Total Nitrogen in Waster Water Based on RBF-NNOptimized by Hierarchical GA [J]. Computer Simulation,2008.25(7):137-140
[57] Spérandio M., and Espinosa M. C., Modelling an aerobic submerged membranebioreactor with ASM models on a large range of sludge retention time [J]. Desalination,2008,231(1-3):82-90
[58] Sotomayor O. A. Z., Park S. W., and Garcia C., Software sensor for on-line estimation ofthe microbial activity in activated sludge systems [J]. ISA Transactions,2002:127-143
[59] Theilliol, D., et al., On-line estimation of unmeasured inputs for anaerobic wastewatertreatment processes [J]. Control Engineering Practice,2003,11(9):1007-1019
[60] Bogaerts P., and Wouwer A. V., Software sensors for bioprocesses [J]. ISA Transactions,2003,42(4):547-558
[61] Alcaraz-González V., et al., Software sensors for highly uncertain WWTPs: a newapproach based on interval observers [J]. Water Research,2002,36(10):2515-2524
[62] Aguado D., et al., A methodology for sequencing batch reactor identification withartificial neural networks: A case study [J]. Computers&Chemical Engineering,2009,33(2):465-472
[63] Zyngier D., et al., Robust soft sensors for SBR monitoring [J]. Water Science andTechnology,2001,43(3):101-105
[64] Feitkenhauer H. and Meyer U., Software sensors based on titrimetric techniques for themonitoring and control of aerobic and anaerobic bioreactors [J]. BiochemicalEngineering Journal,2004,17(2):147-151
[65] Zyngier D., et al., Soft sensors with white-and black-box approaches for a wastewatertreatment process [J]. Brazilian Journal of Chemical Engineering,2000,17:433-440
[66] Zyngier D., et al., Hybrid estimator for a wastewater treatment process [A]. The2ndCongress on Process Systems Engineering for the MERCOSUR FLORIANOPOLIS [C].BRAZIL: LATIN AMERICAN APPLIED RESEARCH,2001:507-511
[67] Onkal Engin, et al., Determination of the relationship between sewage odour and BODby neural networks [J]. Environmental Modelling&Software,2005,20(7):843-850
[68] Qian L., Neural Network Optimization Based on Hierarchical Genetic Algorithm and Itsapplication in Building Soft-sensor for Nitrogen Removal Process in WastewaterTreatment [D]. Shanghai: Department of Control Scicence and Engineering, ShanghaiJiao Tong University,2008
[69] Li Minghe and ming Y., Research and Application of Wastewater TN Soft-sensor Model[J]. Automation&Instrumentation,2009,9:1-4
[70] Pai T. Y., et al., Using fuzzy inference system to improve neural network for predictinghospital wastewater treatment plant effluent [J]. Computers&Chemical Engineering,2009,33(7):1272-1278
[71] Perendeci A., et al., Prediction of effluent quality of an anaerobic treatment plant underunsteady state through ANFIS modeling with on-line input variables [J]. ChemicalEngineering Journal,2008,145(1):78-85
[72] Perendeci A., et al., Effects of phase vector and history extension on prediction power ofadaptive-network based fuzzy inference system (ANFIS) model for a real scale anaerobicwastewater treatment plant operating under unsteady state [J]. Bioresource Technology,2009,100(20):4579-4587
[73] Fortuna L., Graziani S. and Xibilia M. G., Soft sensors for product quality monitoring indebutanizer distillation columns [J]. Control Engineering Practice,2005,13(4):499-508
[74]Li S. J., et al., Soft sensing modeling via artificial neural network based PSO-Alopex [A].Proceedings of2005international conference on machine learning and cybernetics [C].Guangzhou: IEEE,2005:38-43
[75] Alhoniemi E. S. A., Process monitoring and modeling using the self-organizing map [J].Integrated Computer aided Engineering,1999,6(1):3-14
[76] Avoy T. M., Intelligent "control" applications in the process industries [J]. AnnualReviews in Control,2002,26(1):75-86
[77] Ferrer A., et al., PLS: A versatile tool for industrial process improvement andoptimization [J]. Applied Stochastic Models in Business and Industry,2008,24(6):551-567
[78] Rosen C., A Chemometric Approach to Process Monitoring and Control withApplications to Wastewater Treatment Operation [D]. Sweden: Lund Institute ofTechnology,2001
[79] Le Bonté, et al., Toxic event detection by respirometry and adaptive principalcomponents analysis [J]. Environmetrics,2005,16(6):589-601
[80] Lee J. M., Yoo C. and Lee I. B., On-line batch process monitoring using a consecutivelyupdated multiway principal component analysis model [J]. Computers&ChemicalEngineering,2003,27(12):1903-1912
[81] Villez K., Multivariate and qualitative data analysis for monitoring, diagnosis and controlof sequencing batch reactors for wastewater treatment [D]. Belgium: University of Gent,2007
[82] Yoo C. K., Lee D. S. and Vanrolleghem P. A., Application of mutiway ICA for onlineprocess monitoring of a sequencing batch reactor [J]. Water Research,2004,38(7):1715-1732
[83] Durante C., et al., Application of N-PLS to gas chromatographic and sensory data oftraditional balsamic vinegars of modena [J]. Chemometrics and Intelligent LaboratorySystems,2006,83:54-65
[84] Serra P., et al., Development of a real-time expert system for wastewater treatment plantscontrol [J]. Control Engineering Practice,1993,1(2):329-335
[85] Yu R. F., et al., Dynamic control of disinfection for wastewater reuse applying ORP/pHmonitoring and artificial neural networks [J]. Resources, Conservation and Recycling,2008,52(8-9):1015-1021
[86] El-Din A. G. and Smith D. W., A neural network model to predict the wastewater inflowincorporating rainfall events [J]. Water Research,2002,36(5):1115-1126
[87] Berthouex P. M. and Polkowski L. B., Optimum Waste Treatment Plant Design underUncertainty [J]. Water Pollution Control Federation,1970,42(9):1589-1613
[88] Qin X. S., et al, The application of genetic algorithms on grey nonlinear programming inwater environment [J]. Advances in Water Science,2002,13(1):31-36
[89] Xianmin, S. and Weisheng Y., SVI Soft Sensor Design and Modelling [J]. Industry andMine Automation,2007,27(3):27-29
[90] Lumley D., On-line instrument confirmation: how can we check that our instruments areworking?[J]. Water Science and Technology,2002,45(4-5):469-476
[91] Woo S. H., et al., On-line estimation of key process variables based on kernel partialleast squares in an industrial cokes wastewater treatment plant [J]. Journal of HazardousMaterials,2009,161(1):538-544
[92] Zyngier D., et al., Soft sensor development and experimental application to a wastewatertreatment process [J]. Computer Aided Chemical Engineering,8:943-948
[93] Dianlin D., et al. Delay Time Identification and Dynamic Characteristics Study on ANNSoft Sensor [A]. The Sixth International Conference on Intelligent Systems Design andApplications [C]. Jinan:2006:637-642
[94] Warne K., et al., Statistical and computational intelligence techniques for inferentialmodel development: a comparative evaluation and a novel proposition for fusion [J].Engineering Applications of Artificial Intelligence,2004,17(8):871-885
[95] Vijayakumar S. and Schaal S., Incremental Online Learning in High Dimensions [J].Neural Computation,2005,17:2602-2634
[96] Huang M. Z., et al., A fast predicting neural fuzzy model for on-line estimation ofnutrient dynamics in an anoxic/oxic process [J]. Bioresource Technology,2010,101(6):1642-1651
[97] Edakunni N. U., Bayesian Locally Weighted Online Learning [D]. Edinburgh: School ofInformatics, University of Edinburgh,2009
[98] Fujiwara K., Kano M. and Hasebe S., Development of correlation-based patternrecognition and its application to adaptive soft-sensor design [A].2009ICCAS-SICE [C].Fukuoka: IEEE,2009
[99] Fujiwara K., et al., Soft-sensor development using correlation-based just-in-timemodeling [J]. Aiche Journal,2009,55(7):1754-1765
[100] Cheng C. and Chiu M. S., A new data-based methodology for nonlinear processmodeling [J]. Chemical Engineering Science,2004,59(13):2801-2810
[101] Edakunni N. U., Schaal S., and Vijayakumar S., Kernel Carpentry for OnlineRegression using Randomly Varying Coefficient Model [A]. Proceedings of theInternational Joint Conference on Artificial Intelligence [C]. Hyderabad2007
[102] Guang-Bin H., Lei C. and Chee-Kheong S., Universal approximation using incrementalconstructive feedforward networks with random hidden nodes [J]. IEEE Transactionson Neural Networks,200617(4):879-892
[103] Khayamian T., Robustness of PARAFAC and N-PLS regression models in relation tohomoscedastic and heteroscedastic noise [J]. Chemometrics and Intelligent LaboratorySystems,2007,88:35-40
[104] Kadlec P., Gabrys B. and Strandt S., Data-driven Soft Sensors in the process industry[J]. Computers and Chemical Engineering,2009,33(1):795-814
[105] Yue H. H. and Qin S. J., Reconstruction-Based Fault Identification Using a CombinedIndex [J]. Industrial&Engineering Chemistry Research,2001,40(20):4403-4414
[106] Tien D. X., Lim K. W. and Jun L., Comparative study of PCA approaches in processmonitoring and fault detection [A].30th Annual Conference of IEEE in IndustrialElectronics Society [C]. Busan: IEEE,2004
[107] Zhang F. A mixture probabilistic PCA model for multivariate processes monitoring [A].American Control Conference [C]. Washington D. C.: IEEE,2004
[108] Pessel N., et al. An improved PCA fault detection for the diagnosis [A].9th WSEASInternational Conference on Automatic Control, Modeling and Simulation [C]. Istanbul:World Scientific and Engineering Academy and Society,2007
[109] Martin E. B. and Morris A. J., Non-parametric confidence bounds for processperformance monitoring charts [J]. Journal of Process Control,1996,6(6):349-358
[110] Ciba J., et al., PLS-EP algorithm to predict aluminum content in soils of BeskidMountains region [J]. Chemosphere,2009,76(4):565-571
[111] Facco P., et al., Moving average PLS soft sensor for online product quality estimation inan industrial batch polymerization process [J]. Journal of Process Control,2009,19(3):520-529
[112] Cheng C. and Chiu M. S., Nonlinear process monitoring using JITL-PCA [J].Chemometrics and Intelligent Laboratory Systems,2005,76(1):1-13
[113] Merkwirth C., Wichard J. and Ogorzalek M.. A software toolbox for constructingensembles of heterogenous linear and nonlinear models [A]. Proceedings of the2005European Conference on Circuit Theory and Design [C]. Carlo: IEEE,2005
[114] Wichard J. D., Model Selection in an Ensemble Framework [A]. International JointConference on Neural Networks [C]. Vancouver: Computational Intelligent Society,2006
[115] Wichard J. D. and Ogorzalek M.. Time series prediction with ensemble models [A].International Joint Conference on Neural Networks [C]. Vancouver: ComputationalIntelligent Society,2006
[116] Dunia R., et al., Identification of faulty sensors using principal component analysis [J].Aiche Journal,1996,42(10):2797-2812
[117] Papadopoulos H., et al., Inductive Confidence Machines for Regression [A]. ECML2002in Machine Learning [C].2002, Berlin: Springer,185-194
[118] Rallo R. and Giralt F., Best Feature Selection and Data Completion for the Design ofSoft Neural Sensors [A]. Proceedings of AIChE2003,2ndTopical Conference onSensors [C]. San Francisco: ACS,2003
[119] Gernaey K. V., et al., Dynamic influent pollutant disturbance scenario generation usinga phenomenological modelling approach [J]. Environmental Modelling&Software,2011,26(11):1255-1267
[120] Hernebring C, et al., Dynamic online sewer modelling in Helsingborg [J]. WaterScience&Technology,2002,45(4-5):429-436
[121] Langergraber G., et al., Generation of diurnal variation for influent data for dynamicsimulation [J]. Water Science&Technology,2008,57(9):1483-1486
[122] Brockwell P. J. and Davis R. A., Time series: theory and methods [M]. Berlin: SpringerVerlag,2009
[123] Ljung L., System identification [M]. New York: Wiley Online Library,1999
[124] Stoica P. and Nehorai A., On multistep prediction error methods for time series models[J]. Journal of Forecasting,1989,8(4):357-368
[125] Box G. E. P., et al.,, Time series analysis [M]. San Francisco: Holden-day,1970
[126] Isermann R., Parameter adaptive control algorithms-A tutorial [J]. Automatica,1982,18(5):513-528
[127] Sharma K., et al., Predicting hydrogen sulfide formation in sewers: a new model [J].Water,2008,35(2):132-137
[128] Gutierrez O., et al., Effects of long-term pH elevation on the sulfate-reducing andmethanogenic activities of anaerobic sewer biofilms [J]. Water Research,2009,43(9):2549-2557
[129] Ganigué R., et al. Magnesium hydroxide on-line dosing control for sulfide mitigation insewers [A]. Australia's National Water Conference and Exhibtion [C]. Sydney: AWA,2012
[130] Shumway R. H. and Stoffer D. S., Time series analysis and its application [M].3rdEdition ed. Berlin: Springer,2010