基于混合MPLS的多阶段过程质量预报方法

doi:10.6040/j.issn.1672-3961.0.2017.178

山东大学学报(工学版) ›› 2017, Vol. 47 ›› Issue (5): 246-253.doi: 10.6040/j.issn.1672-3961.0.2017.178

基于混合MPLS的多阶段过程质量预报方法

叶晓丰¹, 王培良^1,2*, 杨泽宇¹

1. 杭州电子科技大学新型电子器件与应用研究所, 浙江杭州 310018;2. 湖州师范学院信息与控制技术研究所, 浙江湖州 313000

收稿日期:2017-02-10 出版日期:2017-10-20 发布日期:2017-02-10
通讯作者: 王培良(1963— ),男,浙江湖州人,教授,硕士生导师,主要研究方向为模式识别与智能控制,系统建模与故障诊断,工业自动化等.E-mail:wpl@zjhu.edu.cn E-mail:zhang610378@163.com
作者简介:叶晓丰(1992— ),男,浙江金华人,硕士研究生,主要研究方向为故障诊断与质量预报.E-mail: zhang610378@163.com
基金资助:
国家自然科学基金资助项目(61573137)

Quality prediction method based on hybrid MPLS for multiphases process

YE Xiaofeng¹, WANG Peiliang^1,2*, YANG Zeyu¹

1. Institute of Electron Devices and Application, Hangzhou Dianzi University, Hangzhou 310018, Zhejiang, China;
2. Institute of Information and Control Technology, Huzhou University, Huzhou 313000, Zhejiang, China

Received:2017-02-10 Online:2017-10-20 Published:2017-02-10

摘要/Abstract

摘要： 针对传统的多向偏最小二乘方法(multi-way partial least squares, MPLS)在质量预报中存在着模型预测精度低、局部预报能力不足等问题,提出一种多MPLS模型融合方法来提高预报表现。利用高斯混合模型(Gauss mixture model, GMM)对每批次过程和质量数据组成的高维空间进行阶段识别。针对多批次同一子阶段长度不等问题,采用动态时间规整(dynamic time warping, DTW)算法依据最长持续时间同步为等长轨迹,并在子阶段中按变量展开方式建立MPLS模型。根据Fisher判据分析(Fisher discriminate analysis, FDA)最小化子阶段数据集间相关性,利用核密度方法估计子阶段数据集去相关后的概率密度分布来在线监测阶段切换。利用贝叶斯原则融合各子阶段MPLS模型进行质量预报。将该方法应用到工业青霉素发酵过程中,表明了所提方法具有更好的监控性能和预报能力。

关键词: 多向偏最小二乘方法, 费舍尔判据分析, 多阶段模型融合, 多阶段特性, 贝叶斯原则

Abstract: When the traditional multi-way partial least squares(MPLS)method was carried out for quality prediction, it performed the problems of low prediction accuracy and lack of local capacity. Considering to multiphase characteristics in industrial batch process, a mixture MPLS model method was proposed for the multiphases quality prediction. The Gauss mixture model(GMM)was employed to model the high dimensional spatial distribution of the measurement and the quality variables,which was used to identity sub-phase data blocks in each batch. Due to the problem of inequality length in identical sub-phase, the dynamic time warping(DTW)algorithm corresponding to the maximum path length of time was carried out to synchronize each data block in same sub-phase, whats more, the sub-phase MPLS model was set up according to the variable expansion method. The Fisher discriminate analysis(FDA)was introduced to minimize the relationship among sub-phase blocks, and then the kernel density method was used to monitor phases switch online by estimating the probability density distribution of less relationship sub-phase blocks. The multiphases quality prediction was established by mixing several sub-phase MPLS models according to Bays principle. Furthermore, the result of quality prediction for penicillin fermentation process showed the effectiveness of proposed method.

Key words: multi phase characteristics, bays principle, multi-phase model mixture, multi-way partial least squares, fisher discriminate analysis

中图分类号:

TP277

叶晓丰, 王培良, 杨泽宇. 基于混合MPLS的多阶段过程质量预报方法[J]. 山东大学学报(工学版), 2017, 47(5): 246-253.

YE Xiaofeng, WANG Peiliang, YANG Zeyu. Quality prediction method based on hybrid MPLS for multiphases process[J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2017, 47(5): 246-253.

参考文献

[1] FAN Z, YONG X, ZUO W, et al. Modified principal component analysis: an integration of multiple similarity subspace models[J]. IEEE Transactions on Neural Networks & Learning Systems, 2014, 25(8): 1538-1552.
[2] LYU Z, JIANG Q, YAN X. Batch process monitoring based on multisubspace multiway principal component analysis and time-Series bayesian inference[J]. Industrial & Engineering Chemistry Research, 2014, 53(15): 6457-6466.
[3] WANG B, JIANG Q, YAN X. Fault detection and identification using a Kullback-Leibler divergence based multi-block principal component analysis and bayesian inference [J]. Korean Journal of Chemical Engineering, 2014, 31(6): 930-943.
[4] GODOY J L, VEGA J R, MARCHETTI J L. Relationships between PCA and PLS regression[J]. Chemometrics & Intelligent Laboratory Systems, 2014, 130(2):182-191.
[5] SINGH K P, MALIK A, BASANT N, et al. Multi-way partial least squares modeling of water quality data[J]. Analytica Chimica Acta, 2007, 584(2): 385-396.
[6] DUCHESNE C, MACGREGOR J F. Multivariate analysis and optimization of process variable trajectories for batch processes[J]. Chemometrics & Intelligent Laboratory Systems, 2000, 51(1): 125-137.
[7] CHEN X, GAO X, WANG P, et al. Enhanced fermentation process quality prediction based on multi-phase multi-way partial least squares[J]. Computers & Applied Chemistry, 2011,
[8] 赵春晖, 王福利, 姚远, 等. 基于时段的间歇过程统计建模、在线监测及质量预报[J]. 自动化学报, 2010, 36(3): 366-374. ZHAO C H, WANG F L, YAO Y, et al. Phase-based statistical modeling,online monitoring and quality prediction for batch process[J]. Acta Automatica Sinica, 2010, 36(3):366-374.
[9] ZHAO L P, ZHAO C H, GAO F R. Phase transition analysis based quality prediction for multi-phase batch processes[J]. Chinese Journal of Chemicla Engineering, 2012, 20(6): 1191-1197.
[10] ZHAO C, ZHANG W. Reconstruction based fault diagnosis using concurrent phase partition and analysis of relative changes for multiphase batch processes with limited fault batches[J]. Chemometrics & Intelligent Laboratory Systems, 2014, 130(2):135-150.
[11] 于涛, 王建林, 何坤, 等. 基于MPCA-GP的发酵过程分阶段软测量建模方法 [J]. 仪器仪表学报, 2013, 34(12): 2703-2708. YU T, WANG J L, HE K, et al. Phased soft-sensor modeling method for fermentation process based on MPCA-GP[J].Chinese Journal of Scientific Instrument, 2013, 34(12): 2703-2708.
[12] YU G, LI C, SUN J. Machine fault diagnosis based on Gaussian mixture model and its application [J]. International Journal of Advanced Manufacturing Technology, 2010, 48(1-4): 205-212.
[13] 齐咏生, 王林, 李立, 等. GMM-DPLS间歇过程故障监测与质量预报[J]. 计算机与应用化学, 2013,(10) QI Y S, WANG L, LI L,et al.Fault monitoring and quality prediction for batch process based on GMM-DPLS method[J] , Computers and Application Chemistry. 2013,(10)
[14] CHEN Y, HU B, KEOGH E, et al. DTW-Dtime series semi-supervised learning from a single example[C] // ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. [S.l.] : ACM, 2013:383-391.
[15] DHINGRA S, NIJHAWAN G, PANDIT P. Isolated speech recognition using MFCC and DTW[J]. International Journal of Advanced Research in Electrical Electronics & Instrumentation Engineering, 2013, 2(8):4085-4092.
[16] 陈亚华, 蒋丽英, 郭明,等. 基于多向Fisher判据分析的间歇过程性能监控[J]. 吉林大学学报(信息科学版), 2004, 22(4):384-387. CHEN L H, JIANG L Y, GUO M, et al.Monitoring batch processes using multiway Fisher discriminnant analysis[J].Journal of Jilin University(Information Science Edition), 2004, 22(4):384-387.
[17] YU J W. Rainfall time series forecasting based on modular RBF neural network model coupled with SSA and PLS[J]. Journal of Theoretical & Applied Computer Science, 2012, 6(2): 3-12.
[18] TUO X, LIU M, WANG L, et al. A PLS-Based weighted artificial neural network approach for alpha radioactivity prediction inside contaminated pipes[J]. Mathematical Problems in Engineering, 2014, 2014(1):1-5.
[19] LIU Z, QU J, ZUO M J, et al. Fault level diagnosis for planetary gearboxes using hybrid kernel feature selection and kernel Fisher discriminant analysis[J]. The International Journal of Advanced Manufacturing Technology, 2013, 67(5):1217-1230.
[20] BÍIL M, ANDRÁSIK R, JANOSKA Z. Identification of hazardous road locations of traffic accidents by means of kernel density estimation and cluster significance evaluation[J]. Accident Analysis & Prevention, 2013, 55(3):265-273.
[21] SMITH J, NOURETDINOV I, CRADDOCK R, et al. Anomaly detection of trajectories with kernel density estimation by conformal prediction[J]. IEEE Transactions on Signal Processing, 2014, 437(2):271-280.
[22] GAO Y, KONG X, HU C, et al. Multivariate data modeling using modified kernel partial least squares[J]. Chemical Engineering Research & Design, 2014, 94:466-474.

多维度评价

Viewed

Full text

Abstract

Cited

Shared

Discussed

基于混合MPLS的多阶段过程质量预报方法

Quality prediction method based on hybrid MPLS for multiphases process

PDF (PC)

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 8

多维度评价

本文评价

推荐阅读 0

[1]	程鑫,刘晗,王博,梁典,陈强. 基于双核处理器的主动磁悬浮轴承容错控制架构[J]. 山东大学学报(工学版), 2018, 48(2): 72-80.
[2]	周福娜,高育林,王佳瑜,文成林. 基于深度学习的缓变故障早期诊断及寿命预测[J]. 山东大学学报(工学版), 2017, 47(5): 30-37.
[3]	毛海杰,李炜,王可宏,冯小林. 基于自抗扰的多电机转速同步系统传感器故障切换容错策略[J]. 山东大学学报(工学版), 2017, 47(5): 64-70.
[4]	赵英弘,何潇,周东华. 一类含有传感器故障的网络化系统容错估计[J]. 山东大学学报(工学版), 2017, 47(5): 71-78.
[5]	秦利国,何潇,周东华. 一种时延多智能体系统的分布式编队[J]. 山东大学学报(工学版), 2017, 47(5): 79-88.
[6]	庞人铭,王波,叶昊,张海峰,李明亮. 基于PCA相似度和谱聚类相结合的高炉历史数据聚类[J]. 山东大学学报(工学版), 2017, 47(5): 143-149.
[7]	包塔拉,马剑,甘祖旺. 基于几何特征与流形距离的锂电池健康评估[J]. 山东大学学报(工学版), 2017, 47(5): 157-165.
[8]	王磊,邓晓刚,曹玉苹,田学民. 基于MLFDA的化工过程故障模式分类方法[J]. 山东大学学报(工学版), 2017, 47(5): 179-186.