基于相似性度量的MWSVDD非高斯间歇过程监控
|
摘要
针对间歇过程的非线性、多阶段、过程变量的高斯和非高斯混合分布问题,提出了基于相似性度量的多向加权支持向量数据描述(similarity measure-MWSVDD, SmMWSVDD)算法.该算法首先考虑了阶段间的相似性,将多阶段过程分为稳定阶段和过渡阶段;然后在高维核空间定义了一种新的核相似度权重,对支持向量数据描述(SVDD)建模得到的所有半径进行均衡考虑,克服了SVDD构建控制限的缺陷;通过D检验法将混合分布变量分为高斯分布和非高斯分布变量,并分别用多向核主成分分析(MKPCA)和改进的SVDD进行建模监控;最后通过贝叶斯推断在每个阶段集成各自统一的监控量.通过青霉素发酵实验平台进行验证,结果表明,所提算法比MKPCA和SVDD算法的误报率平均降低了20.21%和漏报率平均降低了10.27%,对多阶段和混合分布的间歇过程监控更加有效.
Aiming at nonlinearity, multiphase and the Gaussian and non-Gaussian mixture distribution of process variables in batch processes, a multiway weighted support vector data description algorithm based on similarity measure MWSVDD(SmMWSVDD) was proposed in this paper. Firstly, the algorithm divided the multiphase process into a stable phase and a transitional phase by considering the similarity between phases. Then, a new kernel similarity weight was defined in high dimensional kernel space to balance all the radiuses obtained by support vector data description(SVDD) modeling, overcoming the shortcoming of the control limits constructed by SVDD. The mixture distribution was divided into Gaussian distribution and non-Gaussian distribution variables by a D-test method to be modeled and monitored using multiway kernel principal component analysis(MKPCA) and improved SVDD. Finally, the integration unified monitoring statistic was built at each phase by Bayesian inference and verified by the penicillin fermentation process. The result shows that the proposed algorithm can reduce the false alarm rate by 20.21% and the missed alarm rate by 10.27% on average than MKPCA and SVDD. Thus, it is more effective for multiphase and mixture distributional batch process monitoring.
Aiming at nonlinearity, multiphase and the Gaussian and non-Gaussian mixture distribution of process variables in batch processes, a multiway weighted support vector data description algorithm based on similarity measure MWSVDD(SmMWSVDD) was proposed in this paper. Firstly, the algorithm divided the multiphase process into a stable phase and a transitional phase by considering the similarity between phases. Then, a new kernel similarity weight was defined in high dimensional kernel space to balance all the radiuses obtained by support vector data description(SVDD) modeling, overcoming the shortcoming of the control limits constructed by SVDD. The mixture distribution was divided into Gaussian distribution and non-Gaussian distribution variables by a D-test method to be modeled and monitored using multiway kernel principal component analysis(MKPCA) and improved SVDD. Finally, the integration unified monitoring statistic was built at each phase by Bayesian inference and verified by the penicillin fermentation process. The result shows that the proposed algorithm can reduce the false alarm rate by 20.21% and the missed alarm rate by 10.27% on average than MKPCA and SVDD. Thus, it is more effective for multiphase and mixture distributional batch process monitoring.
引文
[1] Undey C, Cinar A. Statistical monitoring of multistage, multiphase batch processes[J]. IEEE Control Systems, 2002, 22(5): 40-52. DOI:10.1109/mcs.2002.1035216.
[2] Lv Z, Yan X F, Jiang Q C. Batch process monitoring based on multiple-phase online sorting principal component analysis[J]. ISA Transactions, 2016, 64: 342-352. DOI:10.1016/j.isatra.2016.04.022.
[3] Zhang K, Hao H Y, Chen Z W, et al. A comparison and evaluation of key performance indicator-based multivariate statistics process monitoring approaches[J]. Journal of Process Control, 2015, 33: 112-126. DOI:10.1016/j.jprocont.2015.06.007.
[4] Yin S, Ding S X, Haghani A, et al. A comparison study of basic data-driven fault diagnosis and process monitoring methods on the benchmark Tennessee Eastman process[J]. Journal of Process Control, 2012, 22(9): 1567-1581. DOI:10.1016/j.jprocont.2012.06.009.
[5] Qin S J. Survey on data-driven industrial process monitoring and diagnosis[J]. Annual Reviews in Control, 2012, 36(2): 220-234. DOI:10.1016/j.arcontrol.2012.09.004.
[6] Ge Z Q, Song Z H, Gao F R. Review of recent research on data-based process monitoring[J]. Industrial & Engineering Chemistry Research, 2013, 52(10): 3543-3562. DOI:10.1021/ie302069q.
[7] Zhang Y W. Enhanced statistical analysis of nonlinear processes using KPCA, KICA and SVM[J]. Chemical Engineering Science, 2009, 64(5): 801-811. DOI:10.1016/j.ces.2008.10.012.
[8] Wang H G, Yao M. Fault detection of batch processes based on multivariate functional kernel principal component analysis[J]. Chemometrics and Intelligent Laboratory Systems, 2015, 149: 78-89. DOI:10.1016/j.chemolab.2015.09.018.
[9] Luo L J, Bao S Y, Mao J F, et al. Nonlinear process monitoring based on kernel global-local preserving projections[J]. Journal of Process Control, 2016, 38: 11-21. DOI:10.1016/j.jprocont.2015.12.005.
[10] Luo L J, Bao S Y, Mao J F, et al. Phase partition and phase-based process monitoring methods for multiphase batch processes with uneven durations[J]. Industrial & Engineering Chemistry Research, 2016, 55(7): 2035-2048. DOI:10.1021/acs.iecr.5b03993.
[11] Yu W K, Zhao C H, Zhang S M. A two-step parallel phase partition algorithm for monitoring multiphase batch processes with limited batches[J]. IFAC-PapersOnLine, 2017, 50(1): 2750-2755. DOI:10.1016/j.ifacol.2017.08.582.
[12] Zhao C H. Concurrent phase partition and between-mode statistical analysis for multimode and multiphase batch process monitoring[J]. AIChE Journal, 2014, 60(2): 559-573. DOI:10.1002/aic.14282.
[13] Liu J X, Liu T, Zhang J. Window-based stepwise sequential phase partition for nonlinear batch process monitoring[J]. Industrial & Engineering Chemistry Research, 2016, 55(34): 9229-9243. DOI:10.1021/acs.iecr.6b01257.
[14] Qin Y, Zhao C H, Wang X Z, et al. Subspace decomposition and critical phase selection based cumulative quality analysis for multiphase batch processes[J]. Chemical Engineering Science, 2017, 166: 130-143. DOI:10.1016/j.ces.2017.03.033.
[15] Liu J X, Liu T, Zhang J. Phase partition for nonlinear batch process monitoring[J]. IFAC-PapersOnLine, 2016, 49(7): 1181-1186. DOI:10.1016/j.ifacol.2016.07.367.
[16] Zhao C H, Gao F R. Statistical modeling and online fault detection for multiphase batch processes with analysis of between-phase relative changes[J]. Chemometrics and Intelligent Laboratory Systems, 2014, 130: 58-67. DOI:10.1016/j.chemolab.2013.09.003.
[17] Lv Z, Yan X F, Jiang Q C. Batch process monitoring based on self-adaptive subspace support vector data description[J]. Chemometrics and Intelligent Laboratory Systems, 2017, 170: 25-31. DOI:10.1016/j.chemolab.2017.09.009.
[18] Lv Z, Yan X F. Hierarchical support vector data description for batch process monitoring[J]. Industrial & Engineering Chemistry Research, 2016, 55(34): 9205-9214. DOI:10.1021/acs.iecr.6b00901.
[19] Tax D M J, Duin R P W. Support vector data description[J]. Machine Learning, 2004, 54(1): 45-66. DOI:10.1023/b:mach.0000008084.60811.49.
[20] D'Agostino R B. An omnibus test of normality for moderate and large size samples[J]. Biometrika, 1971, 58(2): 341-348. DOI:10.1093/biomet/58.2.341.
[21] Lee J M, Yoo C, Lee I B. Fault detection of batch processes using multiway kernel principal component analysis[J]. Computers & Chemical Engineering, 2004, 28(9): 1837-1847. DOI:10.1016/j.compchemeng.2004.02.036.
[22] Kourti T. Multivariate dynamic data modeling for analysis and statistical process control of batch processes, start-ups and grade transitions[J]. Journal of Chemometrics, 2003, 17(1): 93-109. DOI:10.1002/cem.778.
[23] Birol G, ündey C, ?inar A. A modular simulation package for fed-batch fermentation: Penicillin production[J]. Computers & Chemical Engineering, 2002, 26(11): 1553-1565. DOI:10.1016/s0098-1354(02)00127-8.
[2] Lv Z, Yan X F, Jiang Q C. Batch process monitoring based on multiple-phase online sorting principal component analysis[J]. ISA Transactions, 2016, 64: 342-352. DOI:10.1016/j.isatra.2016.04.022.
[3] Zhang K, Hao H Y, Chen Z W, et al. A comparison and evaluation of key performance indicator-based multivariate statistics process monitoring approaches[J]. Journal of Process Control, 2015, 33: 112-126. DOI:10.1016/j.jprocont.2015.06.007.
[4] Yin S, Ding S X, Haghani A, et al. A comparison study of basic data-driven fault diagnosis and process monitoring methods on the benchmark Tennessee Eastman process[J]. Journal of Process Control, 2012, 22(9): 1567-1581. DOI:10.1016/j.jprocont.2012.06.009.
[5] Qin S J. Survey on data-driven industrial process monitoring and diagnosis[J]. Annual Reviews in Control, 2012, 36(2): 220-234. DOI:10.1016/j.arcontrol.2012.09.004.
[6] Ge Z Q, Song Z H, Gao F R. Review of recent research on data-based process monitoring[J]. Industrial & Engineering Chemistry Research, 2013, 52(10): 3543-3562. DOI:10.1021/ie302069q.
[7] Zhang Y W. Enhanced statistical analysis of nonlinear processes using KPCA, KICA and SVM[J]. Chemical Engineering Science, 2009, 64(5): 801-811. DOI:10.1016/j.ces.2008.10.012.
[8] Wang H G, Yao M. Fault detection of batch processes based on multivariate functional kernel principal component analysis[J]. Chemometrics and Intelligent Laboratory Systems, 2015, 149: 78-89. DOI:10.1016/j.chemolab.2015.09.018.
[9] Luo L J, Bao S Y, Mao J F, et al. Nonlinear process monitoring based on kernel global-local preserving projections[J]. Journal of Process Control, 2016, 38: 11-21. DOI:10.1016/j.jprocont.2015.12.005.
[10] Luo L J, Bao S Y, Mao J F, et al. Phase partition and phase-based process monitoring methods for multiphase batch processes with uneven durations[J]. Industrial & Engineering Chemistry Research, 2016, 55(7): 2035-2048. DOI:10.1021/acs.iecr.5b03993.
[11] Yu W K, Zhao C H, Zhang S M. A two-step parallel phase partition algorithm for monitoring multiphase batch processes with limited batches[J]. IFAC-PapersOnLine, 2017, 50(1): 2750-2755. DOI:10.1016/j.ifacol.2017.08.582.
[12] Zhao C H. Concurrent phase partition and between-mode statistical analysis for multimode and multiphase batch process monitoring[J]. AIChE Journal, 2014, 60(2): 559-573. DOI:10.1002/aic.14282.
[13] Liu J X, Liu T, Zhang J. Window-based stepwise sequential phase partition for nonlinear batch process monitoring[J]. Industrial & Engineering Chemistry Research, 2016, 55(34): 9229-9243. DOI:10.1021/acs.iecr.6b01257.
[14] Qin Y, Zhao C H, Wang X Z, et al. Subspace decomposition and critical phase selection based cumulative quality analysis for multiphase batch processes[J]. Chemical Engineering Science, 2017, 166: 130-143. DOI:10.1016/j.ces.2017.03.033.
[15] Liu J X, Liu T, Zhang J. Phase partition for nonlinear batch process monitoring[J]. IFAC-PapersOnLine, 2016, 49(7): 1181-1186. DOI:10.1016/j.ifacol.2016.07.367.
[16] Zhao C H, Gao F R. Statistical modeling and online fault detection for multiphase batch processes with analysis of between-phase relative changes[J]. Chemometrics and Intelligent Laboratory Systems, 2014, 130: 58-67. DOI:10.1016/j.chemolab.2013.09.003.
[17] Lv Z, Yan X F, Jiang Q C. Batch process monitoring based on self-adaptive subspace support vector data description[J]. Chemometrics and Intelligent Laboratory Systems, 2017, 170: 25-31. DOI:10.1016/j.chemolab.2017.09.009.
[18] Lv Z, Yan X F. Hierarchical support vector data description for batch process monitoring[J]. Industrial & Engineering Chemistry Research, 2016, 55(34): 9205-9214. DOI:10.1021/acs.iecr.6b00901.
[19] Tax D M J, Duin R P W. Support vector data description[J]. Machine Learning, 2004, 54(1): 45-66. DOI:10.1023/b:mach.0000008084.60811.49.
[20] D'Agostino R B. An omnibus test of normality for moderate and large size samples[J]. Biometrika, 1971, 58(2): 341-348. DOI:10.1093/biomet/58.2.341.
[21] Lee J M, Yoo C, Lee I B. Fault detection of batch processes using multiway kernel principal component analysis[J]. Computers & Chemical Engineering, 2004, 28(9): 1837-1847. DOI:10.1016/j.compchemeng.2004.02.036.
[22] Kourti T. Multivariate dynamic data modeling for analysis and statistical process control of batch processes, start-ups and grade transitions[J]. Journal of Chemometrics, 2003, 17(1): 93-109. DOI:10.1002/cem.778.
[23] Birol G, ündey C, ?inar A. A modular simulation package for fed-batch fermentation: Penicillin production[J]. Computers & Chemical Engineering, 2002, 26(11): 1553-1565. DOI:10.1016/s0098-1354(02)00127-8.