運用雙曲線正切函數監控非線性剖面製程

Using Hyperbolic Tangent Function in Nonlinear Profile Monitoring

李姿儀(Tzu-Yi Lee)

指導教授 ： 范書愷

國立臺北科技大學/管理學院/工業工程與管理系碩士班/碩士(2012年)

https://doi.org/10.6841/NTUT.2012.00458

摘要

在品管應用上，一般而言，製程或產品之品質特性都是針對單一變數進行研究。然而對某些製程而言，品質特性是由反應變數和一個或多個解釋變數間之關係來界定,因此一個品質特性乃以一個函數、一條曲線或是一個曲面之資料型式來呈現，稱之為profile (剖面)。
在本研究提出一個新方法運用在如何有效地監控製程剖面。在此研究中,我們採用參數和非參數方法監控鋁合金於真空爐熱處理過程的溫度曲線。分別以雙曲線正切函數(Hyperbolic tangent)和Smoothing Spline (平滑樣條)來對曲線做配適,模型中參數採用Hotelling’s 統計量和事先制定的測度(metrics)及其管制界線對曲線做監控。在第一階段模擬結果顯示雙曲線正切函數在配適階段表現較平滑樣條佳，在第二階段模擬結果亦顯示雙曲線正切函數在偵測製程異常時的平均連串長度(ARLout)明顯優於平滑樣條。

關鍵字

統計製程管制 ； 非線性剖面 ； 雙曲線正切函數 ； Hotelling’s T2 ； Smoothing Spline(平滑樣條) ； 事先制定的測度(metrics) ； 製程異常下平均連串長度(ARLout)

並列摘要

For most of the SPC applications, the quality of a process or product is measured by one or multiple quality characteristics. Quality characteristics depend on the relationship between the response variable and one and/or explanatory variables. Therefore, a quality characteristic is represented by a function or a curve, which is called for a ‘profile’.
In this thesis, a new method of using the hyperbolic tangent function proposed is for monitoring the profile process. Thus, we adopt the parametric and nonparametric approaches to monitor the vacuum heat treatment process temperature curve. The hyperbolic tangent function is compared to the smoothing spline approach when modeling the nonlinear profiles. The vector of parameter estimates is monitored by Hotelling’s for the parametric approach and metrics method for the nonparametric. The results of the simulation study show that hyperbolic tangent function appears to perform very well for the vacuum heat treatment profiles by Hotelling’s . In Phase I, the proposed hyperbolic tangent approach can correctly identify the outlying profiles but the smoothing spline approach cannot. In phase II, the propose approach provides better out-of-control average run length (ARL) performance than the smoothing spline approach.

並列關鍵字

Nonlinear Profile ； Hyperbolic Tangent Function ； Smoothing spline ； Hotelling’s T2 ； Average Run Length(ARLout)

參考文獻

[3] Y. Ding, L. Zeng and S. Zhou, "Phase I Analysis for Monitoring Nonlinear Profiles in Manufacturing Processes," Journal of Quality Technology, vol. 38, no. 3, 2006, pp. 199-216.

[4] R. Gnanaesikan and J. R. Kettenring, "Robust Estimates, Residuals, and Outlier Detection with Multiresponse Data," Biometrics, vol. 28, no. 1, 1972, pp. 81-124.

[5] S. Gupta, D. C. Montgomery and W. H. Woodall, "Performance evaluation of two methods for online monitoring of linear calibration profiles," International Journal of Production Research, vol. 44, no. 10, 2006, pp. 1927-1942.

[6] D. S. Holmes and A. E. Mergen, "Improving the performance of the Control Chart," Quality Engineering, vol. 4, no. 5, 1993, pp. 619-625.

[7] C. M. Hurvich and C. L. Tsai, "Regression and Time Series Model Selection in Small Samples," Biometrika, vol. 76, no. 2, 1989, pp. 297-307.