Feature Space k Nearest Neighbor Based Batch Process Monitoring
-
摘要: 针对具有非高斯、非线性及多工况特性的批次过程,提出一种基于特征量最近邻统计指标的过程监视方法. 首先,将批次过程正常工况原始数据投影到其特征空间,提取主元T和平方预测误差SPE,并进行特征量k最近邻距离平方和的求解. 然后,采用核密度估计法获得概率密度分布函数,确定统计监视控制限. 特征空间的主元T和SPE特征量能全面代表原始数据的有用信息. 采用特征量k最近邻建立监视模型将会节省存储空间,提高建模样本数量与变量之比以及检测异常工况的速度. 另外,利用局部近邻数据建模可以解决过程具有的非线性和多工况问题,而应用核密度估计法可以解决过程数据具有的非高斯分布问题. 最后,在半导体生产过程的成功应用表明了所提方法的有效性.Abstract: A process monitoring method based on feature space k nearest neighbor is proposed for batch process with the characteristics such as non-Gaussian, nonlinear and multi-mode. Firstly, the normal mode data are projected into the feature space, the principal components T and squared prediction error (SPE) are extracted, and the sum of k nearest neighbor squared distance is computed. Then the kernel density estimation is used to set the statistical threshold of the normal mode. Feature statistic can represent useful information of raw data comprehensively. Modeling using feature statistic can economize storage space, improve the ratio between modeling sample numbers and variables, and increase speed of fault detection. Furthermore, modeling using local neighbor data can solve nonlinear and multi-modes questions of batch process. The use of kernel density estimation method can solve non-Gaussian characteristics of modeling data. The effectiveness of the proposed method is illustrated by applying it to the monitoring of a semiconductor process.
-
[1] Zhou Dong-Hua, Li Gang, Li Yuan. Data-Driven Based Process Fault Detection and Diagnosis Technology. Beijing: Science Press, 2011(周东华, 李钢, 李元. 数据驱动的工业过程故障检测与诊断技术. 北京: 科学出版社, 2011) [2] Lu Ning-Yun, Wang Fu-Li, Gao Fu-Rong, Wang Shu. Statistical modeling and online monitoring for batch process. Acta Automatica Sinica, 2006, 32(3): 400-410(陆宁云, 王福利, 高福荣, 王姝. 间歇过程的统计建模与在线监测. 自动化学报, 2006, 32(3): 400-410) [3] Nomikos P, MacGregor J F. Monitoring batch processes using multiway principal component analysis. American Institute of Chemical Engineers Journal, 1994, 40(8): 1361-1375 [4] Zhao Chun-Hui, Wang Fu-Li, Yao Yuan, Gao Fu-Rong. Phase-based statistical modeling, online monitoring and quality prediction for batch processes. Acta Automatica Sinica, 2010, 36(3): 366-374(赵春晖, 王福利, 姚远, 高福荣. 基于时段的间歇过程统计建模、在线监测及质量预报. 自动化学报, 2010, 36(3): 366-374) [5] Zhao C H, Gao F R. Between-phase-based statistical analysis and modeling for transition monitoring in multiphase batch processes. American Institute of Chemical Engineers Journal, 2012, 58(9): 2682-2696 [6] Qin S J, Cherry G, Good R, Wang J, Harrison C A. Semiconductor manufacturing process control and monitoring: a fab-wide framework. Journal of Process Control, 2006, 16(3): 179-191 [7] Ge Z Q, Song Z H. Semiconductor manufacturing process monitoring based on adaptive substatistical PCA. IEEE Transactions on Semiconductor Manufacturing, 2010, 23(1): 99-108 [8] Ng Y S, Srinivasan R. An adjoined multi-model approach for monitoring batch and transient operations. Computers and Chemical Engineering, 2009, 33(4): 887-902 [9] Chamness K A. Multivariate Fault Detection and Visualization in the Semiconductor Industry[Ph.D. dissertation], The University of Texasat Austin, Austin, 2006 [10] He Q P, Wang J. Fault detection using the k-nearest neighbor rule for semiconductor manufacturing processes. IEEE Transactions on Semiconductor Manufacturing, 2007, 20(4): 345-354 [11] Chamness K, Edgar T F. Local-kNN fault detection algorithm description and examples. In: Proceedings of the 2006 Texas-Wisconsin Modeling and Control Consortium (TWMCC) Conference. Austin, Texas, Spring, 2006. 1-12 [12] He Q P, Wang J. Principal component based k-nearest-neighbor rule for semiconductor process fault detection. In: Proceedings of the 2008 American Control Conference. Seattle, WA: IEEE, 2008. 1606-1611 [13] He Q P, Wang J. Large-scale semiconductor process fault detection using a fast pattern recognition-based method. IEEE Transactions on Semiconductor Manufacturing, 2010, 23(2): 194-200 [14] Verdier G, Ferreira A. Fault detection with an adaptive distance for the k-nearest neighbors rule. In: Proceedings of the 2009 International Conference on Computer and Industrial Engineering. Troyes, France: IEEE, 2009. 1273-1278 [15] Martin E B, Morris A J. Non-parametric confidence bounds for process performance monitoring charts. Journal of Process Control, 1996, 6(6): 349-358 [16] Chen Q, Wynne R J, Goulding P, Sandoz D. The application of principal component analysis and kernel density estimation to enhance process monitoring. Control Engineering Practice, 2000, 8(5): 531-543 [17] Verdier G, Ferreira A. Adaptive mahalanobis distance and k-Nearest Neighbor rule for fault detection in semiconductor manufacturing. IEEE Transactions on Semiconductor Manufacturing, 2011, 24(1): 59-68 [18] Zhang Y W, An J Y, Li Z M, Wang H. Modeling and monitoring for handling nonlinear dynamic processes. Information Sciences, 2013, 235: 97-105 [19] Zhang Y W, Zhou H, Qin S J. Decentralized fault diagnosis of large-scale processes using multiblock kernel principal component analysis. Acta Automatica Sinica, 2010, 36(4): 593-597 [20] Wise B M, Gallagher N B, Butler S W, White D D Jr, Barna G G. A comparison of principal component analysis, multiway principal component analysis, trilinear decomposition and parallel factor analysis for fault detection in a semiconductor etch process. Journal of Chemometrics, 1999, 13(3-4): 379-396 [21] Yu J. Multiway discrete hidden Markov model-based approach for dynamic batch process monitoring and fault classification. American Institute of Chemical Engineers Journal, 2012, 58(9): 2714-2725 [22] 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. Journal of Quality Technology, 2011, 43(2): 142-157 [23] Yu J B. Semiconductor manufacturing process monitoring using Gaussian mixture model and Bayesian method with local and nonlocal information. IEEE Transactions on Semiconductor Manufacturing, 2012, 25(3): 480-493
点击查看大图
计量
- 文章访问数: 1473
- HTML全文浏览量: 77
- PDF下载量: 1044
- 被引次数: 0