We present a new fault detection index, based on Multi-way Principal Component Analysis, which requires no selection of principal and residual spaces. This detection index is called Gaussian Time Error, since a Gaussian model is learned at each measurement instant. This index is used on real data from semiconductor processes to detect faults, providing a better detection than Square Prediction Error in the studied cases.
Gaussian Time Error: A new index for fault detection in semiconductor processes
Rossi, Francesco;
2016-01-01
Abstract
We present a new fault detection index, based on Multi-way Principal Component Analysis, which requires no selection of principal and residual spaces. This detection index is called Gaussian Time Error, since a Gaussian model is learned at each measurement instant. This index is used on real data from semiconductor processes to detect faults, providing a better detection than Square Prediction Error in the studied cases.
File in questo prodotto:
Non ci sono file associati a questo prodotto.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
