全部 标题 作者
关键词 摘要

OALib Journal期刊
ISSN: 2333-9721
费用:99美元

查看量下载量

相关文章

更多...
化工学报  2015 

基于测地线距离统计量的多工况间歇过程监测

DOI: 10.11949/j.issn.0438-1157.20141439, PP. 291-298

Keywords: 多工况,非线性,间歇过程,测地线距离,算法

Full-Text   Cite this paper   Add to My Lib

Abstract:

针对间歇过程数据具有非线性和多工况的特点,提出一种基于测地线距离统计量(geodesicdistancestatistic,GDS)的监测方法。首先,对多工况间歇过程数据按批次方向展开及标准化,利用主元分析(principalcomponentanalysis,PCA)方法进行降维;然后,在降维空间获得赋权邻接矩阵,提出采用改进的Dijkstra(improvedDijkstra,IDijkstra)算法使Dijkstra算法更易于实现,计算各批次之间的测地线距离,用以表征非线性多工况数据之间的实际最短距离,更好地体现批次数据之间的局部近邻关系。通过构造测地线距离α次方统计量Dα进行过程监测,与欧氏距离平方和D2相比将减小边缘训练数据距离的偏离程度。最后,通过在数值仿真和工业仿真实例中的应用,验证所提算法的有效性。

References

[1]  Lee S P, Chao A K, Tsung F, Wong D S H, Tseng S T, Jang S S. Monitoring batch processes with multiple on-off steps in semiconductor manufacturing [J]. Journal of Quality Technology, 2011, 43(2): 142-157
[2]  Wang Shu. A study of batch process fault diagnosis and prediction methods based on data [D]. Shenyang: Northeastern University, 2010
[3]  Chen Yong(陈勇). Fault diagnosis of the production process based on multiway principal components analysis[D]. Hangzhou: Zhejiang University, 2003
[4]  Zhou Donghua(周东华), Li Gang(李钢), Li Yuan(李元). Industrial Process Fault Diagnosis of Data-Driven(数据驱动的工业过程故障诊断技术)[M].Beijing: Science Press, 2011
[5]  Wang Haiqing(王海清), Song Zhihuan(宋执环), Wang Hui(王慧). Fault detection behavior analysis based on PCA process monitoring [J]. Journal of Chemical Industry and Engineering (China)(化工学报), 2002(3): 297-301
[6]  Nomikos P, MacGregor J F. Monitoring batch processes using multiway principal components analysis [J]. AIChE Journal, 1994, 40(8): 1361-1375
[7]  Scholk opf B, Smola A, Muller K R. Nolinear component analysis as a kernel eigenvalue problem [J]. Neural Computation, 1998, 10(5): 1299-1319
[8]  Guo Xiaoping(郭小萍), Yuan Jie(袁杰), Li Yuan(李元). Nearest neighbors index of kernel features based batch processes monitoring and simulation research [J]. Journal of System Simulation(系统仿真学报), 2014, 26(7): 1424-1429
[9]  He Q P, Wang J. Fault detection using the k-nearest neighbor rule for semiconductor manufacturing processes [J]. IEEE Transactions on Semiconductor Manufacturing, 2007, 20(4): 345-354
[10]  Guo Xiaoping(郭小萍), Yuan Jie(袁杰), Li Yuan(李元). Bach process monitoring based on k-nearest neighbor of spatial [J].Journal of Automation(自动化学报),2014,40(1): 135-142
[11]  Bai Zhengguo(白正国), Shen Yibing(沈一兵), Shui Naixiang(水乃翔), et al. Riemannian Introduction(黎曼几何引论)[M]. Beijing: Higher Education Press, 2004
[12]  Jurgen Jost. Riemannian Geometry and Geometric Analysis[M]. Berlin: Springer-Verlag, 2002
[13]  Quan Yong, Yang Jie. Geodesic distance for support vector machines [J]. Acta Automatic Sinica, 2005, 31(2): 202-208
[14]  Yang Qing(杨庆), Chen Guiming(陈桂明), Liu Jingjie(刘鲭洁), He Qingfei(何庆飞). Kernel principal components analysis based on geodesic distance and its application in gear pump fault recognition [J]. Journal of Shanghai Jiao Tong University(上海交通大学学报), 2011, 45(11):1632-1637
[15]  Wu Xiaoting(吴晓婷). Data dimensionality reduction algorithm research based on manifold learning[D]. Dalian: Liaoning Normal University, 2010
[16]  Zhang Fuhao(张福浩), Liu Jiping(刘记平), Li Qingyuan(李青元). Shortest path optimization algorithm based on Dijkstra algorithm [J]. Remote Sensing Information(遥感信息), 2004, 2: 38-41
[17]  Li Chenbo(李臣波), Liu Rentao(刘润涛). A shortest path algorithm based on Dijkstra [J]. Journal of Harbin University of Science and Technology(哈尔滨理工大学学报), 2008, 13(1): 35-37
[18]  Su Baoli(苏宝丽), Li Ning(李宁). The optimization algorithm of Dijkstra and its application for the best path in GIS system [J]. Remote Sensing Technology and Application (遥感技术与应用), 2013, 28 (5):866-870
[19]  He Q P, Wang J. Principal component based k-nearest-neighbor rule for semiconductor process fault detection//Proceedings of the 2008 American Control Conference[C]. Seattle, WA: IEEE, 2008: 1606-1611
[20]  Yu J. Multiway discrete hidden Markov model-based approach for dynamic batch process monitoring and fault classification [J]. American Institute of Chemical Engineers Journal, 2012, 58(9): 2714-2725

Full-Text

Contact Us

service@oalib.com

QQ:3279437679

WhatsApp +8615387084133