基于数据驱动的多模型软测量研究
|
摘要
工业生产过程是一个复杂的过程,存在着多工况、非线性、高噪声等特点。在对其生产过程中的难测变量软测量建模时,如果采用单模型一般难以描述整个工况特性,并对噪声的处理能力比较弱。因此采用多模型软测量的方法能比单模型更好的描述生产过程,使得软测量模型的精度得到提高,鲁棒性得到加强。由于多模型软测量方法有这些优点,本文开展了如下四种不同多模型方法的研究,并将多模型软测量方法应用到一个实际的化工生产装置中。具体的研究成果如下:
1、提出了一种基于加权模糊聚类方法的多模型建模方法。将输入向量与输出的相关性作为加权系数,构建加权模糊聚类算法,对样本空间的输入数据进行聚类,然后分别建立各子模型。测试时采用开关切换方式将输入变量送入对应的子模型进行输出估计,子模型输出作为系统模型的最终输出。该方法能够实现对输入数据更加合理的划分,软测量模型的精度得到了提高。将该方法应用于裂解反应器出口双酚A组分的软测量建模中,仿真结果表明,该方法是一种可行的、有效的软测量建模方法。
2、提出了一种基于疏密部数据划分的软测量多模型建模方法。该方法充分应用了全局核函数和局部核函数的特性,以最近邻聚类法为基础,将输入样本数据分为疏部与多个密部,对疏部采用全局核函数,对密部采用局部核函数,构建加强型支持向量分类机子模型,得到由多模型组成的软测量模型。通过对苯酚蒸发器出口双酚A组分的仿真研究表明,该模型的泛化能力得到了提高。
3、提出了一种基于混沌差分进化模糊C-均值聚类的多模型建模方法。该方法采用混沌差分进化算法对模糊C-均值聚类的目标函数进行全局寻优,能有效的解决模糊C-均值聚类陷入局部最优的问题。将该方法应用于重排反应器出口双酚A组分的软测量建模中,仿真结果表明了该算法构造的软测量多模型的有效性。
4、提出了一种基于改进的局部保持投影算法的多模型建模方法。该方法通过有监督的自适应权值的局部保持投影算法对输入数据空间进行特征提取,并结合最近邻分类器算法进行输入空间的划分,最后融合支持向量机实现多模型建模。将该方法应用于苯酚蒸发器出口双酚A组分的软测量建模中,仿真结果表明,该方法的分类精度高于传统的局部保持投影和最近邻分类器结合算法,该方法的模型估计精度得到了提高,具有更强的泛化能力。
Industrial production is a complex process with characters of multiple-conditions, nonlinear, high noise, etc. When soft approach is used for unpredictable variables, normally single model can’t effectively describe the characteristics of working conditions and is less able to deal with noise. Contrary with single model, multiple models can describe complex production process better and improve the estimated accuracy and generalization ability. According to the engineering application background, four methods for multi-model modeling are proposed in this thesis. Specific results are as follows:
1. A multi-model modeling method based on weighted fuzzy clustering is presented. By using the correlation of input and output as weighted coefficient of fuzzy clustering, it is employed to cluster the input data of sample space, and respectively build sub-models. And then using switch mode, the corresponding sub-models estimate output and final output is determined by the corresponding sub-model output. This method can achieve a more rational division of the input data to improve the accuracy of soft-sensor model. The multi-model is applied to a soft sensor for components of BPA in a cracking reactor exports, and the simulation results show its feasibility and effectiveness.
2. A novel method of multi-modeling for soft sensor is proposed in the paper. The method divides the input sample set into one sparse part and several dense parts based on the nearest neighbor clustering algorithm which it fully apply the characteristics of global and local kernel function. Meanwhile, global kernel and local kernel are applied to construct corresponding sub-model with Enhanced Support Vector Classifier. Finally, a soft sensor system with multi-model is obtained. Using the proposed algorithm to the soft-sensor model of BPA component in a Phenol evaporator outlet, the result of simulation shows the effectiveness of the algorithm.
3. A novel fuzzy C-Mean clustering based on Chaotic Differential Evolution is presented, which is for multiple models soft-sensing modeling. The proposed algorithm optimizes objection function of fuzzy C-Mean clustering by using Chaotic Differential Evolution and gets a global optimal solution, which can effectively address the problems of local optimum for fuzzy C-Mean clustering. The multi-model is applied to estimating the components of BPA in a Rearrange reactor exports, it is shown that the algorithm is effective.
4. A Multi-model soft senor based on Improved Locality Preserving Projection is proposed. The proposed approach extracts the features of input sample space by Supervised and Adapt Weighted Locality Preserving Projection. The multi-models can be constructed by Support Vector Machine after using the nearest neighbor classifier to divide input data space. Using the proposed algorithm to the soft-sensor model of BPA component, the result of simulation shows that the proposed approach has better performance compared with conventional Locality Preserving Projection’s, and has superior in accurate estimation, and generalization ability.
1、提出了一种基于加权模糊聚类方法的多模型建模方法。将输入向量与输出的相关性作为加权系数,构建加权模糊聚类算法,对样本空间的输入数据进行聚类,然后分别建立各子模型。测试时采用开关切换方式将输入变量送入对应的子模型进行输出估计,子模型输出作为系统模型的最终输出。该方法能够实现对输入数据更加合理的划分,软测量模型的精度得到了提高。将该方法应用于裂解反应器出口双酚A组分的软测量建模中,仿真结果表明,该方法是一种可行的、有效的软测量建模方法。
2、提出了一种基于疏密部数据划分的软测量多模型建模方法。该方法充分应用了全局核函数和局部核函数的特性,以最近邻聚类法为基础,将输入样本数据分为疏部与多个密部,对疏部采用全局核函数,对密部采用局部核函数,构建加强型支持向量分类机子模型,得到由多模型组成的软测量模型。通过对苯酚蒸发器出口双酚A组分的仿真研究表明,该模型的泛化能力得到了提高。
3、提出了一种基于混沌差分进化模糊C-均值聚类的多模型建模方法。该方法采用混沌差分进化算法对模糊C-均值聚类的目标函数进行全局寻优,能有效的解决模糊C-均值聚类陷入局部最优的问题。将该方法应用于重排反应器出口双酚A组分的软测量建模中,仿真结果表明了该算法构造的软测量多模型的有效性。
4、提出了一种基于改进的局部保持投影算法的多模型建模方法。该方法通过有监督的自适应权值的局部保持投影算法对输入数据空间进行特征提取,并结合最近邻分类器算法进行输入空间的划分,最后融合支持向量机实现多模型建模。将该方法应用于苯酚蒸发器出口双酚A组分的软测量建模中,仿真结果表明,该方法的分类精度高于传统的局部保持投影和最近邻分类器结合算法,该方法的模型估计精度得到了提高,具有更强的泛化能力。
Industrial production is a complex process with characters of multiple-conditions, nonlinear, high noise, etc. When soft approach is used for unpredictable variables, normally single model can’t effectively describe the characteristics of working conditions and is less able to deal with noise. Contrary with single model, multiple models can describe complex production process better and improve the estimated accuracy and generalization ability. According to the engineering application background, four methods for multi-model modeling are proposed in this thesis. Specific results are as follows:
1. A multi-model modeling method based on weighted fuzzy clustering is presented. By using the correlation of input and output as weighted coefficient of fuzzy clustering, it is employed to cluster the input data of sample space, and respectively build sub-models. And then using switch mode, the corresponding sub-models estimate output and final output is determined by the corresponding sub-model output. This method can achieve a more rational division of the input data to improve the accuracy of soft-sensor model. The multi-model is applied to a soft sensor for components of BPA in a cracking reactor exports, and the simulation results show its feasibility and effectiveness.
2. A novel method of multi-modeling for soft sensor is proposed in the paper. The method divides the input sample set into one sparse part and several dense parts based on the nearest neighbor clustering algorithm which it fully apply the characteristics of global and local kernel function. Meanwhile, global kernel and local kernel are applied to construct corresponding sub-model with Enhanced Support Vector Classifier. Finally, a soft sensor system with multi-model is obtained. Using the proposed algorithm to the soft-sensor model of BPA component in a Phenol evaporator outlet, the result of simulation shows the effectiveness of the algorithm.
3. A novel fuzzy C-Mean clustering based on Chaotic Differential Evolution is presented, which is for multiple models soft-sensing modeling. The proposed algorithm optimizes objection function of fuzzy C-Mean clustering by using Chaotic Differential Evolution and gets a global optimal solution, which can effectively address the problems of local optimum for fuzzy C-Mean clustering. The multi-model is applied to estimating the components of BPA in a Rearrange reactor exports, it is shown that the algorithm is effective.
4. A Multi-model soft senor based on Improved Locality Preserving Projection is proposed. The proposed approach extracts the features of input sample space by Supervised and Adapt Weighted Locality Preserving Projection. The multi-models can be constructed by Support Vector Machine after using the nearest neighbor classifier to divide input data space. Using the proposed algorithm to the soft-sensor model of BPA component, the result of simulation shows that the proposed approach has better performance compared with conventional Locality Preserving Projection’s, and has superior in accurate estimation, and generalization ability.
引文
1.潘丽登,李大宇,马俊英.软测量技术原理与应用[M].北京:中国电力出版社,2008
2.俞金寿.软测量技术及其应用[J].自动化仪表,2008,29(1):1-7
3.黄凤良.软测量思想与软测量技术[J].计量学报,2004,25(3):284-288
4.俞金寿,刘爱伦,张克进.软测量技术及其在石油化工中的应用[M].北京:化学工业出版社,2000
5. Bates J M,Granger C W J.The Combination of Forecasts[J].Operations Research Quarterly,1969,20:319-323
6. Johansen T A,Foss B A.Multiple model approaches to modeling and control [J].Int J Control,1999,72(7/8):575
7.张勇,陈莉.聚类与PCA融合的特征提取方法研究[J].计算机工程与应用,2010,46(11):148-150
8.赵忠盖,刘飞.基于因子回归模型的软测量方法[J].计算机与应用化学,2010,27(1):38-40
9.贾润达,毛志忠,常玉清,等.基于投影寻踪的非线性鲁棒偏最小二乘法及应用[J].控制理论与应用,2010,27(3):391-394
10.吴水秀,曾庆鹏,王明文.基于改进ReliefF算法的主成分特征提取方法[J].计算机工程,2008,34(18):51-52
11.张沐光,宋执环.独立元子空间算法及其在故障检测上的应用[J].化工学报,2010,61(2):425-431
12.刘建敏,刘艳斌,乔新勇,等.基于模糊聚类与神经网络的柴油机技术状态评价方法研究[J].内燃机学报,2008,26(4):379-383
13.王宏伟,于双和.基于目标函数的模糊模型一体化建模[J].控制理论与应用,2010,27(4):523-526
14.李订芳,胡文超,何炎祥.基于共享最近邻聚类和模糊集理论的分类器[J].控制与决策,2006,21(10):1103-1108
15.刘琼荪,杜会锋.基于构造映射的支持向量分类机[J].计算机工程与应用,2009,45(27):130-132
16.袁健,周燕,吕欣.基于潜在成分的时变系统损伤的概率神经网络分类[J].南京航空航天大学学报(英文版),2009,26(4):259-267
17.李俊林,符红光.改进的基于核密度估计的数据分类算法[J].控制与决策,2010,25(4):507-514
18.汪洪桥,孙富春,蔡艳宁,等.多核学习方法[J].自动化学报,2010,36(8):1037-1050
19.呼文亮,王惠文.基于贝叶斯准则的支持向量机预测模型[J].北京航空航天大学学报,2010:36(4):486-489
20.皋军,王士同,邓赵红.基于全局和局部保持的半监督支持向量机[J].电子学报,2010,38(7):1628-1633
21.冯瑞,张玥杰,张艳珠,等.基于加权支持向量机的移动建模方法及其在软测量中的应用[J].自动化学报,2004,30(3):436-441
22. Bezdek J C.Pattern recognition with fuzzy objective function algorithms[M].New York:Plenum Press,1981
23. H?ppner F,Klawonn F,Kruse R,Runkler T.Fuzzy Cluster Analysis.Chichester:John Wiley&Sons Ltd.,1999
24. Vladimir N, Vapnik V. An Overview of Statistical Learning Theory, IEEE Transaction on Neural Networks, 1999, 10(5): 988-999
25. V. Vapnik.统计学习理论的本质[M].第2版.张学工译.北京:清华大学出版社,2000
26. SUN Zong-hai, SUN You-xian. Fuzzy Support Vector Machine for Regression Estimation[C].// Proceedings of The IEEE International Conference on Systems, Man and Cybernetics, 2003, 4:3336-3341
27. SHI Hui-feng,XING Mi-lan.The Improved MC(su3) Algorithm and Bayesian Network Learning[C]//Proceedings of The IEEE International Conference on Machine Learning and Cybernetics, 2008, 3:1775-1778
28.黄金杰,李士勇,蔡云泽.一种建立粗糙数据模型的监督模糊聚类方法[J].软件学报,2005,16(5):744-753
29.皋军,王士同.具有特征排序功能的鲁棒性模糊聚类方法[J].自动化学报, 2009,35(2):145-153
30.陈金凤,杨慧中.混合核支持向量机在化工软测量中的应用研究[J].化工自动化及仪表,2008,35(2):36-38
31.夏红霞,丁子春,李哲,等.一种新的自适应组合核函数[J].武汉理工大学学报,2009,31(3):49-53
32.胡正平,张晔,刘明.分解子空间自适应核函数综合支持向量机算法[J].哈尔滨工业大学学报,2007 39(7):1099-1101
33.李蓉,叶世伟,史忠植. SVM-KNN分类器——一种提高SVM分类精度的新方法[J].电子学报,2002,30(5): 745-748
34.文传军,詹永照.基于自调节分类面SVM的平衡不平衡数据分类[J].系统工程,2009,37(3): 110-113
35.仲蔚,俞金寿.基于模糊C均值聚类的多模型软测量[J].华东理工大学学报,2000,26(1):83-87
36.李卫,杨煜普,王娜.基于核模糊聚类的多模型LSSVM回归建模[J].控制与决策,2008,23(5): 560-562
37.薛振框,李少远.MIMO非线性系统的多模型建模方法[J].电子学报,2005,33(1):52-56
38.徐海霞,刘国海,周大为,等.基于改进核模糊聚类算法的软测量建模研究[J].仪器仪表学报,2009,30(10):2226-2231
39.卢有麟,周建中,李英海,等.混沌差分文化算法及其仿真应用研究[J].系统仿真学报,2009,21(16):5107-5111
40.陈金凤,杨慧中.基于CPSO的混合核SVM参数选择及其应用[J].控制工程,2009,16(1):70-72
41.李超顺,周建中,方仍存,等.基于混沌优化的模糊聚类分析方法[J].系统仿真学报,2009,21(10):2977-2980
42.谭跃,谭冠政.具有混沌局部搜索策略的差分进化全局优化算法[J].计算机工程与应用,2009,45(14):14-17
43.牛大鹏,王福利,何大阔,等.多目标混沌差分进化算法[J].控制与决策,2009,24(3):362-364
44.宗朝霞,汤宏胜,贺曼,等.基于遗传算法的支持向量机预测含能材料密度的研究[J].计算机与应用化学, 2009,26(12):1529-1532
45. He X F, Yan S C, Hu Y, Niyogi P, Zhang H J. Face recognition using Laplacianfaces. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2005, 27(3): 328-340
46.申中华,潘永惠,王士同.有监督的局部保留投影降维算法[J].模式识别与人工智能, 2008, 21(2): 232-239
47.张志伟,杨帆,夏克文,等.一种有监督的LPP算法及其在人脸识别中的应用[J].电子与信息学报, 2008, 30(3): 539-541
48.高全学,谢德燕,徐辉,等.融合局部结构和差异信息的监督特征提取算法[J].自动化学报, 2010, 36(8): 1107-1114
49.林玉娥,顾国昌,刘海波,等.适用于小样本问题的具有类内保持的正交特征提取算法[J].自动化学报, 2010, 36(5): 644-649
50.吕志军,杨建国,项前,等.基于支持向量机的纺纱质量预测模型研究[J].控制与决策, 2007, 22(6): 693-696
2.俞金寿.软测量技术及其应用[J].自动化仪表,2008,29(1):1-7
3.黄凤良.软测量思想与软测量技术[J].计量学报,2004,25(3):284-288
4.俞金寿,刘爱伦,张克进.软测量技术及其在石油化工中的应用[M].北京:化学工业出版社,2000
5. Bates J M,Granger C W J.The Combination of Forecasts[J].Operations Research Quarterly,1969,20:319-323
6. Johansen T A,Foss B A.Multiple model approaches to modeling and control [J].Int J Control,1999,72(7/8):575
7.张勇,陈莉.聚类与PCA融合的特征提取方法研究[J].计算机工程与应用,2010,46(11):148-150
8.赵忠盖,刘飞.基于因子回归模型的软测量方法[J].计算机与应用化学,2010,27(1):38-40
9.贾润达,毛志忠,常玉清,等.基于投影寻踪的非线性鲁棒偏最小二乘法及应用[J].控制理论与应用,2010,27(3):391-394
10.吴水秀,曾庆鹏,王明文.基于改进ReliefF算法的主成分特征提取方法[J].计算机工程,2008,34(18):51-52
11.张沐光,宋执环.独立元子空间算法及其在故障检测上的应用[J].化工学报,2010,61(2):425-431
12.刘建敏,刘艳斌,乔新勇,等.基于模糊聚类与神经网络的柴油机技术状态评价方法研究[J].内燃机学报,2008,26(4):379-383
13.王宏伟,于双和.基于目标函数的模糊模型一体化建模[J].控制理论与应用,2010,27(4):523-526
14.李订芳,胡文超,何炎祥.基于共享最近邻聚类和模糊集理论的分类器[J].控制与决策,2006,21(10):1103-1108
15.刘琼荪,杜会锋.基于构造映射的支持向量分类机[J].计算机工程与应用,2009,45(27):130-132
16.袁健,周燕,吕欣.基于潜在成分的时变系统损伤的概率神经网络分类[J].南京航空航天大学学报(英文版),2009,26(4):259-267
17.李俊林,符红光.改进的基于核密度估计的数据分类算法[J].控制与决策,2010,25(4):507-514
18.汪洪桥,孙富春,蔡艳宁,等.多核学习方法[J].自动化学报,2010,36(8):1037-1050
19.呼文亮,王惠文.基于贝叶斯准则的支持向量机预测模型[J].北京航空航天大学学报,2010:36(4):486-489
20.皋军,王士同,邓赵红.基于全局和局部保持的半监督支持向量机[J].电子学报,2010,38(7):1628-1633
21.冯瑞,张玥杰,张艳珠,等.基于加权支持向量机的移动建模方法及其在软测量中的应用[J].自动化学报,2004,30(3):436-441
22. Bezdek J C.Pattern recognition with fuzzy objective function algorithms[M].New York:Plenum Press,1981
23. H?ppner F,Klawonn F,Kruse R,Runkler T.Fuzzy Cluster Analysis.Chichester:John Wiley&Sons Ltd.,1999
24. Vladimir N, Vapnik V. An Overview of Statistical Learning Theory, IEEE Transaction on Neural Networks, 1999, 10(5): 988-999
25. V. Vapnik.统计学习理论的本质[M].第2版.张学工译.北京:清华大学出版社,2000
26. SUN Zong-hai, SUN You-xian. Fuzzy Support Vector Machine for Regression Estimation[C].// Proceedings of The IEEE International Conference on Systems, Man and Cybernetics, 2003, 4:3336-3341
27. SHI Hui-feng,XING Mi-lan.The Improved MC(su3) Algorithm and Bayesian Network Learning[C]//Proceedings of The IEEE International Conference on Machine Learning and Cybernetics, 2008, 3:1775-1778
28.黄金杰,李士勇,蔡云泽.一种建立粗糙数据模型的监督模糊聚类方法[J].软件学报,2005,16(5):744-753
29.皋军,王士同.具有特征排序功能的鲁棒性模糊聚类方法[J].自动化学报, 2009,35(2):145-153
30.陈金凤,杨慧中.混合核支持向量机在化工软测量中的应用研究[J].化工自动化及仪表,2008,35(2):36-38
31.夏红霞,丁子春,李哲,等.一种新的自适应组合核函数[J].武汉理工大学学报,2009,31(3):49-53
32.胡正平,张晔,刘明.分解子空间自适应核函数综合支持向量机算法[J].哈尔滨工业大学学报,2007 39(7):1099-1101
33.李蓉,叶世伟,史忠植. SVM-KNN分类器——一种提高SVM分类精度的新方法[J].电子学报,2002,30(5): 745-748
34.文传军,詹永照.基于自调节分类面SVM的平衡不平衡数据分类[J].系统工程,2009,37(3): 110-113
35.仲蔚,俞金寿.基于模糊C均值聚类的多模型软测量[J].华东理工大学学报,2000,26(1):83-87
36.李卫,杨煜普,王娜.基于核模糊聚类的多模型LSSVM回归建模[J].控制与决策,2008,23(5): 560-562
37.薛振框,李少远.MIMO非线性系统的多模型建模方法[J].电子学报,2005,33(1):52-56
38.徐海霞,刘国海,周大为,等.基于改进核模糊聚类算法的软测量建模研究[J].仪器仪表学报,2009,30(10):2226-2231
39.卢有麟,周建中,李英海,等.混沌差分文化算法及其仿真应用研究[J].系统仿真学报,2009,21(16):5107-5111
40.陈金凤,杨慧中.基于CPSO的混合核SVM参数选择及其应用[J].控制工程,2009,16(1):70-72
41.李超顺,周建中,方仍存,等.基于混沌优化的模糊聚类分析方法[J].系统仿真学报,2009,21(10):2977-2980
42.谭跃,谭冠政.具有混沌局部搜索策略的差分进化全局优化算法[J].计算机工程与应用,2009,45(14):14-17
43.牛大鹏,王福利,何大阔,等.多目标混沌差分进化算法[J].控制与决策,2009,24(3):362-364
44.宗朝霞,汤宏胜,贺曼,等.基于遗传算法的支持向量机预测含能材料密度的研究[J].计算机与应用化学, 2009,26(12):1529-1532
45. He X F, Yan S C, Hu Y, Niyogi P, Zhang H J. Face recognition using Laplacianfaces. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2005, 27(3): 328-340
46.申中华,潘永惠,王士同.有监督的局部保留投影降维算法[J].模式识别与人工智能, 2008, 21(2): 232-239
47.张志伟,杨帆,夏克文,等.一种有监督的LPP算法及其在人脸识别中的应用[J].电子与信息学报, 2008, 30(3): 539-541
48.高全学,谢德燕,徐辉,等.融合局部结构和差异信息的监督特征提取算法[J].自动化学报, 2010, 36(8): 1107-1114
49.林玉娥,顾国昌,刘海波,等.适用于小样本问题的具有类内保持的正交特征提取算法[J].自动化学报, 2010, 36(5): 644-649
50.吕志军,杨建国,项前,等.基于支持向量机的纺纱质量预测模型研究[J].控制与决策, 2007, 22(6): 693-696