傅永峰,徐欧官,陈祥华,陈伟杰

• 1:浙江工业大学之江学院

摘要(Abstract)：

针对目前软测量建模过程中,单个模型难以精确描述复杂非线性对象而多模型又多采用静态模型因而对系统实际运行中的动态变化考虑不足的问题,提出了一种基于多模型动态融合的自适应软测量建模方法。该方法首先采用仿射传播聚类算法对样本数据进行分类,并对不同类别的输入样本分别建立基于高斯过程回归的子模型,最后使用动态Gauss-Markov估计对各子模型估计值进行融合。将上述方法应用于对二甲苯(p-xylene,简称PX)吸附分离过程纯度的软测量建模,仿真结果表明该方法能够有效地增强模型适应工况变化的能力,是一种有效的软测量建模方法。

关键词(KeyWords)： 软测量;仿射传播(AP);高斯过程回归(GPR);动态Gauss-Markov估计;PX纯度

Abstract:

Keywords:

基金项目(Foundation): 国家自然科学基金(61304125,61203133)

作者(Author): 傅永峰,徐欧官,陈祥华,陈伟杰

参考文献(References)：

• [1]Mc Avoy T J.Contemplative stance for chemical process control[J].Automatica,1992,28(2):441-442.
• [2]Cao P F,Luo X L.Modeling of soft sensor for chemical process[J].J of Chem Ind&Eng,2013,64(3):88-800.
• [3]Hector J G,He Q P,Wang J.Comparison of the performance of a reduced-order dynamic PLS soft sensor with different updating schemes for digester control[J].Contr Eng Pract,2012,20(8):747-760.
• [4]Liu J L.Developing a soft sensor based on sparse partial least squares with variable selection[J].J of Process Contr,2014,24(7):1046-1056.
• [5]Liu J L.On-line soft sensor for polyethylene process with multiple production grades[J].Contr Eng Pract,2007,15(7):769-778.
• [6]Bates J M,Granger C W J.The combination of forecasts[J].Oper Res,1969,20(4):451-468.
• [7]FU Yong-feng(傅永峰).Adaptive soft-sensor modeling algorithm based on FCMISVM and its application in PX adsorption separation process(基于FCMISVM的自适应软测量建模方法及其在PX吸附分离过程中的应用)[J].Chinese J of Chem Eng(化工学报英文版),2008,16(5):746-751.
• [8]LI Xiu-liang(李修亮),SU Hong-ye(苏宏业),CHU Jian(褚健).Multiple model soft sensor based on affinity propagation,Gaussian process and Bayesian committee machine(基于仿射传播,高斯过程和贝叶斯决策的多模型软测量)[J].Chinese J of Chem Eng(化工学报英文版),2009,17(1):95-99.
• [9]LI Li-juan(李丽娟),SONG Kun(宋坤),ZHAO Ying-kai(赵英凯).Modeling of ARA fermentation based on affinity propagation clustering(基于仿射传播聚类的ARA发酵过程建模)[J].J of Chem Ind&Eng(化工学报),2011,62(8):2116-2121.
• [10]LI Ya-qin(李雅琴),YANG Hui-zhong(杨慧中).Multi-model modeling method based on affinity propagation clustering and Gaussian Processes(基于仿射传播聚类和高斯过程的多模型建模方法)[J].Comput&Appl Chem(计算机与应用化学),2010,27(1):51-54.
• [11]LI Wei(李炜),ZHANG Yin(章寅).Research and application of soft sensor method based on heterogeneous multi-model dynamic fusion(异类多模型动态融合软测量方法研究与应用)[J].Contr&Instrum in Chem Ind(化工自动化及仪表),2010,37(8):42-45.
• [12]YU Tao(于涛),WANG Jian-lin(王建林),HE Kun(何坤),et al.Staged soft sensor modeling method for fermentation process based on MPCA-GP(基于MPCA-GP的发酵过程分阶段软测量建模方法)[J].Chinese J of Sci Instrum(仪器仪表学报),2013,34(12):2703-2708.
• [13]Frey J,Dueck D.Clustering by passing messages between data points[J].Sci,2007,315(5814):972-976.
• [14]Rasmussen C E,Williams C K I.Gaussian process for machine learning[M].Cambridge:MIT Press,2006:8-31.
• [15]Likar B,Kocijan J.Predictive control of a gas-liquid separation plant based on Gaussian process model[J].Comput&Chem Eng,2007,31(3):142-152.
• [16]Pavone D,Hotier G.System approach modeling applied to the eluxyl process[J].Oil Gas Sci&Tech,2000,55(4):437-446.
• [17]Minceva M,Rodrigues A E.Modeling and simulation of a simulated moving bed for the separation of p-xylene[J].Ind&Eng Chem Res,2002,41(14):3454-3461.
• [18]XU Ou-guan(徐欧官),CHEN Xiang-hua(陈祥华).Recursive PLS soft-sensor with moving window of fixed length and its commercial application(移动窗递推PLS软测量建模及其工业应用)[J].J Chem Eng of Chinese Univ(高校化学工程学报),2009,23(6):1044-1050.