A Multivariate Moving Average Control Chart for Photovoltaic Processes

Authors: Chunchom Pongchavalit

Abstract:

For the electrical metrics that describe photovoltaic cell performance are inherently multivariate in nature, use of a univariate, or one variable, statistical process control chart can have important limitations. Development of a comprehensive process control strategy is known to be significantly beneficial to reducing process variability that ultimately drives up the manufacturing cost photovoltaic cells. The multivariate moving average or MMA chart, is applied to the electrical metrics of photovoltaic cells to illustrate the improved sensitivity on process variability this method of control charting offers. The result show the ability of the MMA chart to expand to as any variables as needed, suggests an application with multiple photovoltaic electrical metrics being used in concert to determine the processes state of control.

Keywords: The multivariate moving average control chart, Photovoltaic processes control, Multivariate system.

Digital Object Identifier (DOI): doi.org/10.5281/zenodo.1072359

References:

[1] Coleman J. (1999). Process Optimization and Control of a Photovoltaic Manufacturing Process. Extended Abstracts and Papers 9th Workshop on Crystalline Sillicon Solar Cell Material and Processes, 171-174.
[2] Goulding, P.R.,et al. (2000). Fault detection in continuous processes using multivariate statistical methods. International Journal of Systems Science 31(11), 1459-1471.
[3] Lucas, J.M.,&Saccucci, M.S.(1990). Exponential weight moving average control schemes: Properties and enhancements Technometrics, 32.
[4] McCafferty, R.H.(2001). High road to process control: Multivariate methods Semiconductor International, 24(8), 257-261.
[5] Minitab Inc. (2003). Statguide Minitab Statistical Software v14.State College PA.
[6] Montgomery D, (2001). Introduction to Statistical Quality Control, Fourth Editor, John Wiley, NY.
[7] Prabhu, S.S.,&Runger, G.C. (1997). Designing a multivariate EWMA control chart. Journal of Quality Technology, 26.
[8] Skinner, K.R.et al. (2002). Multivariate statistical methods for modeling and analysis of wafer probe test data. IEEE Transactions on Semiconductor Manufacturing, 15(4), 523-530.
[9] Tseng, S.T., Chou, R.J.&Leee, S.P. (2002). A study on amultivariate EWMA controller. IIE Transactions 34, 541-549.