基于EGK’M-RBF神经网络的软测量建模与强化学习控制算法的研究
|
摘要
论文一方面从解决顺丁橡胶聚合过程中门尼粘度在线测量问题出发,提出采用软测量技术来建立门尼粘度软仪表。通过对顺丁橡胶聚合过程中的工艺流程进行分析研究后,提出一种基于EGK'M-RBF网络的软测量建模方法,并利用顺丁橡胶聚合过程现场数据,建立了基于EGK'M-RBF网络的门尼粘度软测量模型。
另一方面从解决复杂过程的控制问题出发,提出采用强化学习理论来进行控制。通过对酿酒酵母发酵过程以及强化学习控制理论进行研究,提出一种改进的多步行为Q学习控制算法,并将该算法应用到酿酒酵母发酵过程控制中。论文的主要研究内容和研究成果包括：
1.首先对软测量和强化学习技术进行了概述,接着介绍了基于RBF网络的软测量技术的基本理论,并分析了RBF网络中心选取算法的优缺点以及采用PCA和KPCA进行非线性特征提取实现辅助变量选择的优缺点。
2.通过分析,提出一种增强的全局K'-means聚类算法,并将该算法成功应用于高斯数据集和实际数据集的分类过程中。通过与改进的全局K-means算法以及K'-means算法进行比较,实际数据的实验结果证明所提出的算法能获得更好的聚类结果。接着将增强的全局K'-means聚类算法成功应用于RBF网络隐含层结构的确定,最后提出一种基于EGK'M-RBF网络的软测量建模方法,该方法采用本文提出的EGK'M来确定RBF网络隐含层的结构,采用KPCA实现辅助变量的二次选择,并给出基于EGK'M-RBF神经网络的在线自校正模型算法步骤。
3.提出采用EGK'M-RBF神经网络建立顺丁橡胶门尼粘度的软测量模型,通过对模型结果的比较分析,可知EGK'M-RBF神经网络在门尼粘度软测量建模中具有优势。与其他三种模型相比,本文所提出的模型具有更好的拟合效果,更强的预测能力,更小的泛化误差。同时还对PCA和KPCA进行非线性特征提取进行了比较分析,可知KPCA更适合非线性特征提取。最后还简单介绍了顺丁橡胶门尼粘度软测量软件包的开发以及界面的设计。
4.通过对强化学习控制算法以及酿酒酵母发酵过程进行研究,提出一种改进的多步行为Q学习控制算法,该算法由多步行为Q学习算法和一个模糊控制增益参数选择器组成。通过这种模糊控制增益参数选择器来自适应地选择控制增益参数,一方面加快控制跟踪的速度,另一方面有助于减少控制器的超调。实验结果证明,改进的多步行为Q学习控制器具有超调量小、跟踪速度快,过渡时间短,控制作用平稳等特点。
Firstly, to solve the problem of on-line measurement of mooney viscosity of polybutadiene rubber, a soft sensor of mooney viscosity was proposed by using soft-sensing technology. Through analysis and researches on the process of polybutadiene rubber, an modeling method by radial basis function(RBF) network based on enhanced global K'-means algorithm(EGK'M) was presented, a soft-sensor model of mooney viscosity based on EGK'M-RBF network was established using field data of the process of polybutadiene rubber
Secondly, in order to tackle the control problem of complex industrial process, a method for complex industrial process control using reinforcement learning algorithm was proposed. Through research on the process of Saccharomyces cerevisiae fermentation and reinforcement learning algorithm control theory, an improved multi-step action Q-learning control algorithm is presented. Algorithm was developed to control the ethanol concentration of the Saccharomyces cerevisiae fermentation process. Main contributions of the thesis are as follows:
1. At the beginning, overviews are made both on soft sensing technology and reinforcement learning technology. Then, principle theory of soft sensing based on RBF network was introduced, including the advantages and disadvantages of RBF network center selection algorithms, the advantages and disadvantages of PCA and KPCA for extracting non-linear feature information and achieving selection of auxiliary variable.
2. An enhanced global K'-means clustering algorithm is presented, and it had been developed for clustering Gaussian datasets and several actual datasets. The clustering results of the actual datasets demonstrate that the enhanced global K'-means algorithm can get better clustering results compared to the modified global K-means algorithm and K'-means algorithm, respectively. Then, the enhanced global K'-means algorithm was applied to determine the structure of the hidden layer of RBF network. A modeling method by radial basis function (RBF) network based on enhanced global K'-means algorithm (EGK'M) was presented. In the proposed method the structure of RBF network was detrmined through EGK'M algorithm, KPCA algorithm was used for non-linear feature information and secondary selection of auxiliary variable. Finally, in the thesis, a series procedure of on-line self-calibration model algorithm which based on EGK'M-RBF network was given.
3. Modeling of mooney viscosity of polybutadiene rubber with EGK'M-RBF network was proposed. From the comparative analysis of the modeling results, one can see that advantages of EGK'M-RBF network lies in that the model proposed much better fitting results, stronger predictive ability, smaller absolute error. At the same time, comparative analysis on PCA and KPCA for extracting non-linear feature information shows that KPCA is more suitable for non-linear feature extraction. Finally, a brief introduction to the development and interface design of soft-sensing software package of Mooney viscosity of polybutadiene rubber was given.
4. An improved multi-step action Q-learning control algorithm was proposed for the process of Saccharomyces cerevisiae fermentation, which combines multi-step action Q-learning algorithm and a fuzzy control gain parameter selector. The fuzzy control gain parameter selector was used to adaptively select the control gain parameter, it can lead to faster tracking and help to alleviate the overshoot of controller. Experiment results show that the improved multi-step action Q-learning controller has much lower overshoot, faster tracking, shorter transition, and smoother control signal and so on.
另一方面从解决复杂过程的控制问题出发,提出采用强化学习理论来进行控制。通过对酿酒酵母发酵过程以及强化学习控制理论进行研究,提出一种改进的多步行为Q学习控制算法,并将该算法应用到酿酒酵母发酵过程控制中。论文的主要研究内容和研究成果包括：
1.首先对软测量和强化学习技术进行了概述,接着介绍了基于RBF网络的软测量技术的基本理论,并分析了RBF网络中心选取算法的优缺点以及采用PCA和KPCA进行非线性特征提取实现辅助变量选择的优缺点。
2.通过分析,提出一种增强的全局K'-means聚类算法,并将该算法成功应用于高斯数据集和实际数据集的分类过程中。通过与改进的全局K-means算法以及K'-means算法进行比较,实际数据的实验结果证明所提出的算法能获得更好的聚类结果。接着将增强的全局K'-means聚类算法成功应用于RBF网络隐含层结构的确定,最后提出一种基于EGK'M-RBF网络的软测量建模方法,该方法采用本文提出的EGK'M来确定RBF网络隐含层的结构,采用KPCA实现辅助变量的二次选择,并给出基于EGK'M-RBF神经网络的在线自校正模型算法步骤。
3.提出采用EGK'M-RBF神经网络建立顺丁橡胶门尼粘度的软测量模型,通过对模型结果的比较分析,可知EGK'M-RBF神经网络在门尼粘度软测量建模中具有优势。与其他三种模型相比,本文所提出的模型具有更好的拟合效果,更强的预测能力,更小的泛化误差。同时还对PCA和KPCA进行非线性特征提取进行了比较分析,可知KPCA更适合非线性特征提取。最后还简单介绍了顺丁橡胶门尼粘度软测量软件包的开发以及界面的设计。
4.通过对强化学习控制算法以及酿酒酵母发酵过程进行研究,提出一种改进的多步行为Q学习控制算法,该算法由多步行为Q学习算法和一个模糊控制增益参数选择器组成。通过这种模糊控制增益参数选择器来自适应地选择控制增益参数,一方面加快控制跟踪的速度,另一方面有助于减少控制器的超调。实验结果证明,改进的多步行为Q学习控制器具有超调量小、跟踪速度快,过渡时间短,控制作用平稳等特点。
Firstly, to solve the problem of on-line measurement of mooney viscosity of polybutadiene rubber, a soft sensor of mooney viscosity was proposed by using soft-sensing technology. Through analysis and researches on the process of polybutadiene rubber, an modeling method by radial basis function(RBF) network based on enhanced global K'-means algorithm(EGK'M) was presented, a soft-sensor model of mooney viscosity based on EGK'M-RBF network was established using field data of the process of polybutadiene rubber
Secondly, in order to tackle the control problem of complex industrial process, a method for complex industrial process control using reinforcement learning algorithm was proposed. Through research on the process of Saccharomyces cerevisiae fermentation and reinforcement learning algorithm control theory, an improved multi-step action Q-learning control algorithm is presented. Algorithm was developed to control the ethanol concentration of the Saccharomyces cerevisiae fermentation process. Main contributions of the thesis are as follows:
1. At the beginning, overviews are made both on soft sensing technology and reinforcement learning technology. Then, principle theory of soft sensing based on RBF network was introduced, including the advantages and disadvantages of RBF network center selection algorithms, the advantages and disadvantages of PCA and KPCA for extracting non-linear feature information and achieving selection of auxiliary variable.
2. An enhanced global K'-means clustering algorithm is presented, and it had been developed for clustering Gaussian datasets and several actual datasets. The clustering results of the actual datasets demonstrate that the enhanced global K'-means algorithm can get better clustering results compared to the modified global K-means algorithm and K'-means algorithm, respectively. Then, the enhanced global K'-means algorithm was applied to determine the structure of the hidden layer of RBF network. A modeling method by radial basis function (RBF) network based on enhanced global K'-means algorithm (EGK'M) was presented. In the proposed method the structure of RBF network was detrmined through EGK'M algorithm, KPCA algorithm was used for non-linear feature information and secondary selection of auxiliary variable. Finally, in the thesis, a series procedure of on-line self-calibration model algorithm which based on EGK'M-RBF network was given.
3. Modeling of mooney viscosity of polybutadiene rubber with EGK'M-RBF network was proposed. From the comparative analysis of the modeling results, one can see that advantages of EGK'M-RBF network lies in that the model proposed much better fitting results, stronger predictive ability, smaller absolute error. At the same time, comparative analysis on PCA and KPCA for extracting non-linear feature information shows that KPCA is more suitable for non-linear feature extraction. Finally, a brief introduction to the development and interface design of soft-sensing software package of Mooney viscosity of polybutadiene rubber was given.
4. An improved multi-step action Q-learning control algorithm was proposed for the process of Saccharomyces cerevisiae fermentation, which combines multi-step action Q-learning algorithm and a fuzzy control gain parameter selector. The fuzzy control gain parameter selector was used to adaptively select the control gain parameter, it can lead to faster tracking and help to alleviate the overshoot of controller. Experiment results show that the improved multi-step action Q-learning controller has much lower overshoot, faster tracking, shorter transition, and smoother control signal and so on.
引文
[1]曹柳林,江弘,陈红.利用神经元网络实现门尼粘度的预估[J].北京化工大学学报,1997,24(4)：60-64
[2]张磊,胡春,钱锋.BP改进算法及其在乙二醇精制软测量中的应用[J].自动化仪表,2005,26(006)：31-34
[3]Azlan Hussain M. Review of the applications of neural networks in chemical process control-Simulation and online implementation[J]. Artificial intelligence in engineering, 1999,13(1):55-68
[4]陈兰星.基于RBF神经元网络的聚酯工业软测量软件包的研究及开发[D]. 北京化工大学,2005
[5]俞佩菲.阿维菌素发酵过程中的软测量技术应用研究[D].浙江大学,2006
[6]孔建益,李公法,熊禾根,蒋国璋,杨金堂,王兴东,侯宇.工业生产中软测量建模方法及其应用研究[J].机床与液压,2007,35(006)：149-151
[7]Weber R., Brosilow C. The use of secondary measurements to improve control[J]. AIChE Journal,1972,18(3):614-623
[8]Joseph B., Brosilow C., Howell J., Kerr W. Multi-temps give better control[J]. Hydrocarbon Processing,1976,55(3):127-131
[9]Brosilow C. Inferential control of process control[J]. AIChe J,1978,24(3):475-484
[10]刘颖.电站锅炉风粉浓度的软测量研究[D].东南大学,2006
[11]Tham M., Montague G., Julian Morris A., Lant P. Soft-sensors for process estimation and inferential control[J]. Journal of Process Control,1991,1(1):3-14
[12]Quintero-Marmol E., Luyben W., Georgakis C. Application of an extended Luenberger observer to the control of multicomponent batch distillation[J]. Industrial & Engineering Chemistry Research,1991,30(8):1870-1880
[13]Guilandoust M., Morris A., Tham M. An adaptive estimation algorithm for inferential control[J]. Industrial & Engineering Chemistry Research,1988,27(9):1658-1664
[14]李海青,黄志尧.软测量技术原理及应用[J].化学工业出版社,北京,2000
[15]Kosanovich K., Dahl K., Piovoso M. Improved process understanding using multiway principal component analysis[J]. Ind. Eng. Chem. Res,1996,35(1):138-146
[16]陈守煜,陈晓冰.模糊模式识别理论模型及在化工中的应用[J].化工学报,1991,42(006)：666-668
[17]Babu ka R., Verbruggen H. An overview of fuzzy modeling for control[J]. Control Engineering Practice,1996,4(11):1593-1606
[18]Zhang J., Morris A. Fuzzy neural networks for nonlinear systems modelling[J]. IEE Proceedings-Control Theory and Applications,1995,142-551
[19]Sarimveis H., Alexandridis A., Tsekouras G., Bafas G A fast and efficient algorithm for training radial basis function neural networks based on a fuzzy partition of the input space[J]. Ind. Eng. Chem. Res,2002,41(4):751-759
[20]Arauzo-Bravo M., Cano-Izquierdo J., Gomez-Sanchez E., Lopez-Nieto M., Dimitriadis Y., Lopez-Coronado J. Automatization of a penicillin production process with soft sensors and an adaptive controller based on neuro fuzzy systems[J]. Control Engineering Practice, 2004,12(9):1073-1090
[21]Willis M., Montague G., Di Massimo C., Tham M., Morris A. Artificial neural networks in process estimation and control[J]. Automatica (Journal of IFAC),1992,28(6): 1181-1187
[22]Psichogios D., Ungar L. Direct and indirect model based control using artificial neural networks[J]. Industrial & Engineering Chemistry Research,1991,30(12):2564-2573
[23]闭治跃,王庆丰,唐建中.基于径向基函数神经网络的挖泥船排泥管道泥浆浓度软测量模型研究[J].传感技术学报,2007,20(007)：1630-1634
[24]Silva F., Almeida L. Speeding up backpropagation[J]. Advanced neural computers,1990, 92:151-160
[25]王旭东,邵惠鹤.改进的RBF神经元网络及其应用[J].上海交通大学学报,1996,30(004)：132-136
[26]Wang S., Yu D. Adaptive RBF network for parameter estimation and stable air-fuel ratio control[J]. Neural Networks,2008,21(1):102-112
[27]刘俊,杨春节,卢建刚,陈金水,孙优贤.改进RBF网络设计方法及其在软测量建模上的应用[J].江南大学学报：自然科学版,2009,8(002)：135-139
[28]俞金寿.软测量新技术综述[J].世界仪表与自动化,2001,5(009)：16-22
[29]袁曾任.人工神经元网络及其应用[M].清华大学出版社：广西科学技术出版社,1999
[30]颜学峰,余娟,钱锋.基于径基函数-偏最小二乘回归的对羧基苯甲醛含量软测量模型[J].石油炼制与化工,2005,36(012)：50-53
[31]Padmavathi G., Mandan M., Mitra S., Chaudhuri K. Neural modelling of Mooney viscosity of polybutadiene rubber[J]. Computers and Chemical Engineering,2005,29(7): 1677-1685
[32]陶斌军.基于DRNN的聚酯粘度软仪表的研究[D].北京化工大学,2001
[33]刘载文,王正祥,王小艺,杨斌,程志强.过程神经元网络学习算法及软测量方法的研究[J].系统仿真学报,2007,19(007)：1456-1459
[34]骆中华.基于数据驱动的软测量建模技术及其工业应用[D].浙江大学,2006
[35]罗荣富,邵惠鹤.推断控制中二次变量选择方法的研究[J].1992年中国控制与决策学术年会论文集,1992(6)：146-149
[36]刘良宏,周兴贵.非线性分布参数系统状态估计的最佳测量位置[J].化工学报,1996,47(003)：267-272
[37]Zamprogna E., Barolo M., Seborg D. Optimal selection of soft sensor inputs for batch distillation columns using principal component analysis[J]. Journal of Process Control, 2005,15(1):39-52
[38]马朝阳,苏宏业,傅永峰,褚健.基于KPCA-SVR方法的复合肥养分含量建模[J].中国科学技术大学学报,2005,35(增刊)：314-321
[39]Lee J., Yoo C., Choi S., Vanrolleghem P., Lee I. Nonlinear process monitoring using kernel principal component analysis[J]. Chemical engineering science,2004,59(1): 223-234
[40]刘瑞兰,陈渭泉,苏宏业.基于改进GA-PLS算法的最优辅助变量选择及其在软测量建模中的应用[J].南京邮电大学学报：自然科学版,2006,26(001)：76-80
[41]Tong H., Crowe C. Detection of gross errors in data reconciliation by principal component analysis[J].AIChE Journal,1995,41(7):1712-1722
[42]王芳,岳金彩,谭心舜,郑世清.用于过失测量数据侦破与校正的改进MT-NT算法[J].计算机与应用化学,2002,19(5)：562-566
[43]童豪.生物发酵过程中的软测量技术应用研究[D].浙江大学,2004
[44]鄂加强,王耀南,梅炽.铜精炼过程铜液温度软测量模型及应用[J].化工学报,2006,57(001)：203-209
[45]Lan Z., Nan Y., Yong-zai L. Modelling and control for nonlinear time-delay system via pattern recognition approach[J]. Annual Review in Automatic Programming,1989, 15(Part 2):43-48
[46]傅永峰.软测量建模方法研究及其工业应用[D].浙江大学,2007
[47]Minsky M. Theory of neural-analog reinforcement systems and its application to the brain-model problem[M]. Princeton University,1954
[48]Barto A., Sutton R., Brouwer P. Associative search network:A reinforcement learning associative memory[J]. Biological cybernetics,1981,40(3):201-211
[49]Sutton R. Temporal credit assignment in reinforcement learning[D]. Electronic Doctoral Dissertations for UMass Amherst,1984
[50]Sutton R. Learning to predict by the methods of temporal differences[J]. Machine learning,1988,3(1):9-44
[51]Watkins C. Learning from delayed rewards[D]. King's College, University of Cambridge, 1989
[52]Watkins C., Dayan P. Q-learning[J]. Machine learning,1992,8(3):279-292
[53]Waltz M., Fu K. A heuristic approach to reinforcement learning control systems[J]. IEEE Transactions on Automatic Control,1965,10(4):390-398
[54]Saridis G. Self-organizing control of stochastic systems[M]. M. Dekker,1977
[55]Barto A., Sutton R., Anderson C. Neuronlike adaptive elements that can solve difficult learning control problems[J]. IEEE Transactions on systems, man, and cybernetics,1983, 13(5):834-846
[56]Singh S., Sutton R. Reinforcement learning with replacing eligibility traces[J]. Machine learning,1996,22(1):123-158
[57]Ningshou X., Zhanglei W., Liping C. A learning modified generalized predictive controller[J]. IFAC Symposium on Intelligent Tuning and Adaptive Control,1991: 231-236
[58]阎平凡.再励学习——原理,算法及其在智能控制中的应用[J].信息与控制,1996,25(001)：28-34
[59]Syafiie S., Tadeo F., Martinez E. Model-free learning control of neutralization processes using reinforcement learning[J]. Engineering Applications of Artificial Intelligence,2007, 20(6):767-782
[60]Dantigny P. Modeling of the aerobic growth of Saccharomyces cerevisiae on mixtures of glucose and ethanol in continuous culture[J]. Journal of biotechnology,1995,43(3): 213-220
[61]Lee C., Nakano A., Shiomi N., Lee E., Katoh S. Effects of substrate feed rates on heterologous protein expression by Pichia pastoris in DO-stat fed-batch fermentation[J]. Enzyme and Microbial Technology,2003,33(4):358-365
[62]Tan T., Zhang M., Gao H. Ergosterol production by fed-batch fermentation of Saccharomyces cerevisiae[J]. Enzyme and Microbial Technology,2003,33(4):366-370
[63]Lee J., Lee S., Park S., Middelberg A. Control of fed-batch fermentations [J]. Biotechnology advances,1999,17(1):29-48
[64]Nagy Z. Model based control of a yeast fermentation bioreactor using optimally designed artificial neural networks[J]. Chemical Engineering Journal,2007,127(1-3):95-109
[65]awrynczuk M. Modelling and nonlinear predictive control of a yeast fermentation biochemical reactor using neural networks[J]. Chemical Engineering Journal,2008, 145(2):290-307
[66]Xiong Z., Zhang J. Modelling and optimal control of fed-batch processes using a novel control affine feedforward neural network[J]. Neurocomputing,2004,61:317-337
[67]Gadkar K., Mehra S., Gomes J. On-line adaptation of neural networks for bioprocess control [J]. Computers and Chemical Engineering,2005,29(5):1047-1057
[68]Mitsche D. Review of "Introduction to clustering large and high-dimensional data" by J. Kogan[J]. Computer Science Review,2008,2(1):60-62
[69]Lu J. F., Tang J. B., Tang Z. M., Yang J. Y. Hierarchical initialization approach for K-Means clustering[J]. Pattern Recognition Letters,2008,29(6):787-795
[70]Zalik K. R. An efficient k'-means clustering algorithm[J]. Pattern Recognition Letters, 2008,29(9):1385-1391
[71]Jain A., Murty M., Flynn P. Data clustering:a review[J]. ACM computing surveys,1999, 31(3):264-323
[72]陈衡岳.聚类分析及聚类结果评估算法研究[D].沈阳：东北大学,2006
[73]Fisher D. Knowledge acquisition via incremental conceptual clustering[J]. Machine learning,1987,2(2):139-172
[74]Hinneburg A., Keim D. An efficient approach to clustering in large multimedia databases with noise[J]. Knowledge Discovery and Data Mining,1998,58-65
[75]Kaufman L., Rousseeuw P. Finding groups in data:an introduction to cluster analysis[J]. New York,1990
[76]Akaike H. Information theory and an extension of the maximum likelihood principle[J]. Selected Papers of Hirotugu Akaike,1998:267-281
[77]Schwarz G. Estimating the dimension of a model[J]. The annals of statistics,1978,6(2): 461-464
[78]Wallace C., Dowe D. Minimum message length and Kolmogorov complexity[J]. The Computer Journal,1999,42(4):270-283
[79]Xu L., Krzyzak A., Oja E. Rival penalized competitive learning for clustering analysis, RBFnet, and curve detection[J]. IEEE Transactions on Neural networks,1993,4(4): 636-649
[80]Law L., Cheung Y. Color image segmentation using rival penalized controlled competitive learning[J]. IJCNN 2003,2003
[81]Cheung Y On rival penalization controlled competitive learning for clustering with automatic cluster number selection[J]. IEEE Transactions on Knowledge and Data Engineering,2005,17(11):1583-1588
[82]Ma J., Cao B. The Mahalanobis Distance Based Rival Penalized Competitive Learning Algorithm[J]. Lecture Notes in Computer Science,2006,442-447
[83]Fayyad U., Reina C., Bradley P. Initialization of iterative refinement clustering algorithms[J].1998:194-198
[84]Khan S. S., Ahmad A. Cluster center initialization algorithm for K-means clustering[J]. Pattern Recognition Letters,2004,25(11):1293-1302
[85]Redmond S., Heneghan C. A method for initialising the K-means clustering algorithm using kd-trees[J]. Pattern Recognition Letters,2007,28(8):965-973
[86]Laszlo M., Mukherjee S. A genetic algorithm that exchanges neighboring centers for k-means clustering[J]. Pattern Recognition Letters,2007,28(16):2359-2366
[87]Hansen P., Ngai E., Cheung B., Mladenovic N. Analysis of global k-means, an incremental heuristic for minimum sum-of-squares clustering [J]. Journal of classification, 2005,22(2):287-310
[88]Likas A., Vlassis N., J. Verbeek J. The global k-means clustering algorithm[J]. Pattern Recognition,2003,36(2):451-461
[89]Bagirov A., Yearwood J. A new nonsmooth optimization algorithm for minimum sum-of-squares clustering problems[J]. European Journal of Operational Research,2006, 170(2):578-596
[90]Bagirov A., Mardaneh K. Modified global k-means algorithm for clustering in gene expression data sets[J]. Australian Computer Society, Inc.,2006
[91]Bagirov A. Modified global k-means algorithm for minimum sum-of-squares clustering problems[J]. Pattern Recognition,2008,41(10):3192-3199
[92]UCI repository of machine learning databases, [http://ics. uci. edu/mlearn/MLRepository. html].
[93]Psychol J., Generalis A., Genet S., Biol M., Bioinformatics B., Anal C., Chemom J.1. Fisher R:The use of multiple measurements in taxonomic problems[J]. Ann of Eugenics, 1936,7:179-188
[94]Abdin E., Xu W. Control design and dynamic performance analysis of a windturbine-induction generator unit[J]. IEEE Transaction on Energy Conversion,2000, 15(1):91-96
[95]Schoknecht R., Riedmiller M. Learning to control at multiple time scales[J]. Lecture Notes in Computer Science,2003:479-487
[96]Hauskrecht M., Meuleau N., Kaelbling L., Dean T., Boutilier C. Hierarchical solution of Markov decision processes using macro-actions. Citeseer,1998:220-229
[97]Aha D. A study of instance-based learning algorithms for supervised learning tasks: Mathematical, empirical, and psychological evaluations[D]. PhD Thesis. Technical Report 90-42. University of California at Irvine, Department of Information and Computer Science,1990
[98]Pfluger N., Yen J., Langari R. A defuzzification strategy for a fuzzy logic controller employingprohibitive information in command formulation.1992:717-723
[2]张磊,胡春,钱锋.BP改进算法及其在乙二醇精制软测量中的应用[J].自动化仪表,2005,26(006)：31-34
[3]Azlan Hussain M. Review of the applications of neural networks in chemical process control-Simulation and online implementation[J]. Artificial intelligence in engineering, 1999,13(1):55-68
[4]陈兰星.基于RBF神经元网络的聚酯工业软测量软件包的研究及开发[D]. 北京化工大学,2005
[5]俞佩菲.阿维菌素发酵过程中的软测量技术应用研究[D].浙江大学,2006
[6]孔建益,李公法,熊禾根,蒋国璋,杨金堂,王兴东,侯宇.工业生产中软测量建模方法及其应用研究[J].机床与液压,2007,35(006)：149-151
[7]Weber R., Brosilow C. The use of secondary measurements to improve control[J]. AIChE Journal,1972,18(3):614-623
[8]Joseph B., Brosilow C., Howell J., Kerr W. Multi-temps give better control[J]. Hydrocarbon Processing,1976,55(3):127-131
[9]Brosilow C. Inferential control of process control[J]. AIChe J,1978,24(3):475-484
[10]刘颖.电站锅炉风粉浓度的软测量研究[D].东南大学,2006
[11]Tham M., Montague G., Julian Morris A., Lant P. Soft-sensors for process estimation and inferential control[J]. Journal of Process Control,1991,1(1):3-14
[12]Quintero-Marmol E., Luyben W., Georgakis C. Application of an extended Luenberger observer to the control of multicomponent batch distillation[J]. Industrial & Engineering Chemistry Research,1991,30(8):1870-1880
[13]Guilandoust M., Morris A., Tham M. An adaptive estimation algorithm for inferential control[J]. Industrial & Engineering Chemistry Research,1988,27(9):1658-1664
[14]李海青,黄志尧.软测量技术原理及应用[J].化学工业出版社,北京,2000
[15]Kosanovich K., Dahl K., Piovoso M. Improved process understanding using multiway principal component analysis[J]. Ind. Eng. Chem. Res,1996,35(1):138-146
[16]陈守煜,陈晓冰.模糊模式识别理论模型及在化工中的应用[J].化工学报,1991,42(006)：666-668
[17]Babu ka R., Verbruggen H. An overview of fuzzy modeling for control[J]. Control Engineering Practice,1996,4(11):1593-1606
[18]Zhang J., Morris A. Fuzzy neural networks for nonlinear systems modelling[J]. IEE Proceedings-Control Theory and Applications,1995,142-551
[19]Sarimveis H., Alexandridis A., Tsekouras G., Bafas G A fast and efficient algorithm for training radial basis function neural networks based on a fuzzy partition of the input space[J]. Ind. Eng. Chem. Res,2002,41(4):751-759
[20]Arauzo-Bravo M., Cano-Izquierdo J., Gomez-Sanchez E., Lopez-Nieto M., Dimitriadis Y., Lopez-Coronado J. Automatization of a penicillin production process with soft sensors and an adaptive controller based on neuro fuzzy systems[J]. Control Engineering Practice, 2004,12(9):1073-1090
[21]Willis M., Montague G., Di Massimo C., Tham M., Morris A. Artificial neural networks in process estimation and control[J]. Automatica (Journal of IFAC),1992,28(6): 1181-1187
[22]Psichogios D., Ungar L. Direct and indirect model based control using artificial neural networks[J]. Industrial & Engineering Chemistry Research,1991,30(12):2564-2573
[23]闭治跃,王庆丰,唐建中.基于径向基函数神经网络的挖泥船排泥管道泥浆浓度软测量模型研究[J].传感技术学报,2007,20(007)：1630-1634
[24]Silva F., Almeida L. Speeding up backpropagation[J]. Advanced neural computers,1990, 92:151-160
[25]王旭东,邵惠鹤.改进的RBF神经元网络及其应用[J].上海交通大学学报,1996,30(004)：132-136
[26]Wang S., Yu D. Adaptive RBF network for parameter estimation and stable air-fuel ratio control[J]. Neural Networks,2008,21(1):102-112
[27]刘俊,杨春节,卢建刚,陈金水,孙优贤.改进RBF网络设计方法及其在软测量建模上的应用[J].江南大学学报：自然科学版,2009,8(002)：135-139
[28]俞金寿.软测量新技术综述[J].世界仪表与自动化,2001,5(009)：16-22
[29]袁曾任.人工神经元网络及其应用[M].清华大学出版社：广西科学技术出版社,1999
[30]颜学峰,余娟,钱锋.基于径基函数-偏最小二乘回归的对羧基苯甲醛含量软测量模型[J].石油炼制与化工,2005,36(012)：50-53
[31]Padmavathi G., Mandan M., Mitra S., Chaudhuri K. Neural modelling of Mooney viscosity of polybutadiene rubber[J]. Computers and Chemical Engineering,2005,29(7): 1677-1685
[32]陶斌军.基于DRNN的聚酯粘度软仪表的研究[D].北京化工大学,2001
[33]刘载文,王正祥,王小艺,杨斌,程志强.过程神经元网络学习算法及软测量方法的研究[J].系统仿真学报,2007,19(007)：1456-1459
[34]骆中华.基于数据驱动的软测量建模技术及其工业应用[D].浙江大学,2006
[35]罗荣富,邵惠鹤.推断控制中二次变量选择方法的研究[J].1992年中国控制与决策学术年会论文集,1992(6)：146-149
[36]刘良宏,周兴贵.非线性分布参数系统状态估计的最佳测量位置[J].化工学报,1996,47(003)：267-272
[37]Zamprogna E., Barolo M., Seborg D. Optimal selection of soft sensor inputs for batch distillation columns using principal component analysis[J]. Journal of Process Control, 2005,15(1):39-52
[38]马朝阳,苏宏业,傅永峰,褚健.基于KPCA-SVR方法的复合肥养分含量建模[J].中国科学技术大学学报,2005,35(增刊)：314-321
[39]Lee J., Yoo C., Choi S., Vanrolleghem P., Lee I. Nonlinear process monitoring using kernel principal component analysis[J]. Chemical engineering science,2004,59(1): 223-234
[40]刘瑞兰,陈渭泉,苏宏业.基于改进GA-PLS算法的最优辅助变量选择及其在软测量建模中的应用[J].南京邮电大学学报：自然科学版,2006,26(001)：76-80
[41]Tong H., Crowe C. Detection of gross errors in data reconciliation by principal component analysis[J].AIChE Journal,1995,41(7):1712-1722
[42]王芳,岳金彩,谭心舜,郑世清.用于过失测量数据侦破与校正的改进MT-NT算法[J].计算机与应用化学,2002,19(5)：562-566
[43]童豪.生物发酵过程中的软测量技术应用研究[D].浙江大学,2004
[44]鄂加强,王耀南,梅炽.铜精炼过程铜液温度软测量模型及应用[J].化工学报,2006,57(001)：203-209
[45]Lan Z., Nan Y., Yong-zai L. Modelling and control for nonlinear time-delay system via pattern recognition approach[J]. Annual Review in Automatic Programming,1989, 15(Part 2):43-48
[46]傅永峰.软测量建模方法研究及其工业应用[D].浙江大学,2007
[47]Minsky M. Theory of neural-analog reinforcement systems and its application to the brain-model problem[M]. Princeton University,1954
[48]Barto A., Sutton R., Brouwer P. Associative search network:A reinforcement learning associative memory[J]. Biological cybernetics,1981,40(3):201-211
[49]Sutton R. Temporal credit assignment in reinforcement learning[D]. Electronic Doctoral Dissertations for UMass Amherst,1984
[50]Sutton R. Learning to predict by the methods of temporal differences[J]. Machine learning,1988,3(1):9-44
[51]Watkins C. Learning from delayed rewards[D]. King's College, University of Cambridge, 1989
[52]Watkins C., Dayan P. Q-learning[J]. Machine learning,1992,8(3):279-292
[53]Waltz M., Fu K. A heuristic approach to reinforcement learning control systems[J]. IEEE Transactions on Automatic Control,1965,10(4):390-398
[54]Saridis G. Self-organizing control of stochastic systems[M]. M. Dekker,1977
[55]Barto A., Sutton R., Anderson C. Neuronlike adaptive elements that can solve difficult learning control problems[J]. IEEE Transactions on systems, man, and cybernetics,1983, 13(5):834-846
[56]Singh S., Sutton R. Reinforcement learning with replacing eligibility traces[J]. Machine learning,1996,22(1):123-158
[57]Ningshou X., Zhanglei W., Liping C. A learning modified generalized predictive controller[J]. IFAC Symposium on Intelligent Tuning and Adaptive Control,1991: 231-236
[58]阎平凡.再励学习——原理,算法及其在智能控制中的应用[J].信息与控制,1996,25(001)：28-34
[59]Syafiie S., Tadeo F., Martinez E. Model-free learning control of neutralization processes using reinforcement learning[J]. Engineering Applications of Artificial Intelligence,2007, 20(6):767-782
[60]Dantigny P. Modeling of the aerobic growth of Saccharomyces cerevisiae on mixtures of glucose and ethanol in continuous culture[J]. Journal of biotechnology,1995,43(3): 213-220
[61]Lee C., Nakano A., Shiomi N., Lee E., Katoh S. Effects of substrate feed rates on heterologous protein expression by Pichia pastoris in DO-stat fed-batch fermentation[J]. Enzyme and Microbial Technology,2003,33(4):358-365
[62]Tan T., Zhang M., Gao H. Ergosterol production by fed-batch fermentation of Saccharomyces cerevisiae[J]. Enzyme and Microbial Technology,2003,33(4):366-370
[63]Lee J., Lee S., Park S., Middelberg A. Control of fed-batch fermentations [J]. Biotechnology advances,1999,17(1):29-48
[64]Nagy Z. Model based control of a yeast fermentation bioreactor using optimally designed artificial neural networks[J]. Chemical Engineering Journal,2007,127(1-3):95-109
[65]awrynczuk M. Modelling and nonlinear predictive control of a yeast fermentation biochemical reactor using neural networks[J]. Chemical Engineering Journal,2008, 145(2):290-307
[66]Xiong Z., Zhang J. Modelling and optimal control of fed-batch processes using a novel control affine feedforward neural network[J]. Neurocomputing,2004,61:317-337
[67]Gadkar K., Mehra S., Gomes J. On-line adaptation of neural networks for bioprocess control [J]. Computers and Chemical Engineering,2005,29(5):1047-1057
[68]Mitsche D. Review of "Introduction to clustering large and high-dimensional data" by J. Kogan[J]. Computer Science Review,2008,2(1):60-62
[69]Lu J. F., Tang J. B., Tang Z. M., Yang J. Y. Hierarchical initialization approach for K-Means clustering[J]. Pattern Recognition Letters,2008,29(6):787-795
[70]Zalik K. R. An efficient k'-means clustering algorithm[J]. Pattern Recognition Letters, 2008,29(9):1385-1391
[71]Jain A., Murty M., Flynn P. Data clustering:a review[J]. ACM computing surveys,1999, 31(3):264-323
[72]陈衡岳.聚类分析及聚类结果评估算法研究[D].沈阳：东北大学,2006
[73]Fisher D. Knowledge acquisition via incremental conceptual clustering[J]. Machine learning,1987,2(2):139-172
[74]Hinneburg A., Keim D. An efficient approach to clustering in large multimedia databases with noise[J]. Knowledge Discovery and Data Mining,1998,58-65
[75]Kaufman L., Rousseeuw P. Finding groups in data:an introduction to cluster analysis[J]. New York,1990
[76]Akaike H. Information theory and an extension of the maximum likelihood principle[J]. Selected Papers of Hirotugu Akaike,1998:267-281
[77]Schwarz G. Estimating the dimension of a model[J]. The annals of statistics,1978,6(2): 461-464
[78]Wallace C., Dowe D. Minimum message length and Kolmogorov complexity[J]. The Computer Journal,1999,42(4):270-283
[79]Xu L., Krzyzak A., Oja E. Rival penalized competitive learning for clustering analysis, RBFnet, and curve detection[J]. IEEE Transactions on Neural networks,1993,4(4): 636-649
[80]Law L., Cheung Y. Color image segmentation using rival penalized controlled competitive learning[J]. IJCNN 2003,2003
[81]Cheung Y On rival penalization controlled competitive learning for clustering with automatic cluster number selection[J]. IEEE Transactions on Knowledge and Data Engineering,2005,17(11):1583-1588
[82]Ma J., Cao B. The Mahalanobis Distance Based Rival Penalized Competitive Learning Algorithm[J]. Lecture Notes in Computer Science,2006,442-447
[83]Fayyad U., Reina C., Bradley P. Initialization of iterative refinement clustering algorithms[J].1998:194-198
[84]Khan S. S., Ahmad A. Cluster center initialization algorithm for K-means clustering[J]. Pattern Recognition Letters,2004,25(11):1293-1302
[85]Redmond S., Heneghan C. A method for initialising the K-means clustering algorithm using kd-trees[J]. Pattern Recognition Letters,2007,28(8):965-973
[86]Laszlo M., Mukherjee S. A genetic algorithm that exchanges neighboring centers for k-means clustering[J]. Pattern Recognition Letters,2007,28(16):2359-2366
[87]Hansen P., Ngai E., Cheung B., Mladenovic N. Analysis of global k-means, an incremental heuristic for minimum sum-of-squares clustering [J]. Journal of classification, 2005,22(2):287-310
[88]Likas A., Vlassis N., J. Verbeek J. The global k-means clustering algorithm[J]. Pattern Recognition,2003,36(2):451-461
[89]Bagirov A., Yearwood J. A new nonsmooth optimization algorithm for minimum sum-of-squares clustering problems[J]. European Journal of Operational Research,2006, 170(2):578-596
[90]Bagirov A., Mardaneh K. Modified global k-means algorithm for clustering in gene expression data sets[J]. Australian Computer Society, Inc.,2006
[91]Bagirov A. Modified global k-means algorithm for minimum sum-of-squares clustering problems[J]. Pattern Recognition,2008,41(10):3192-3199
[92]UCI repository of machine learning databases, [http://ics. uci. edu/mlearn/MLRepository. html].
[93]Psychol J., Generalis A., Genet S., Biol M., Bioinformatics B., Anal C., Chemom J.1. Fisher R:The use of multiple measurements in taxonomic problems[J]. Ann of Eugenics, 1936,7:179-188
[94]Abdin E., Xu W. Control design and dynamic performance analysis of a windturbine-induction generator unit[J]. IEEE Transaction on Energy Conversion,2000, 15(1):91-96
[95]Schoknecht R., Riedmiller M. Learning to control at multiple time scales[J]. Lecture Notes in Computer Science,2003:479-487
[96]Hauskrecht M., Meuleau N., Kaelbling L., Dean T., Boutilier C. Hierarchical solution of Markov decision processes using macro-actions. Citeseer,1998:220-229
[97]Aha D. A study of instance-based learning algorithms for supervised learning tasks: Mathematical, empirical, and psychological evaluations[D]. PhD Thesis. Technical Report 90-42. University of California at Irvine, Department of Information and Computer Science,1990
[98]Pfluger N., Yen J., Langari R. A defuzzification strategy for a fuzzy logic controller employingprohibitive information in command formulation.1992:717-723