论文信息 - Fault detection in reactive ion etching systems using one-class support vector machines

Fault detection in reactive ion etching systems using one-class support vector machines

A robust method to detect faults in reactive ion etching systems using optical emission spectroscopy data is proposed. The approach is based on one-class support vector machines (SVMs). Unlike previously proposed fault detection methods, this approach only requires data collected during normal equipment operation to be trained. The results obtained suggest that this technique can detect equipment faults with exceptional accuracy. The SVM used detected all faults, yielding a detection accuracy of 100% with zero false alarms

G.S. May | T. Sarmiento | S.J. Hong

[1] Bernhard Schölkopf,et al. Estimating the Support of a High-Dimensional Distribution , 2001, Neural Computation.

[2] G.S. May,et al. Neural network-based real-time malfunction diagnosis of reactive ion etching using in situ metrology data , 2004, IEEE Transactions on Semiconductor Manufacturing.

[3] Nello Cristianini,et al. An introduction to Support Vector Machines , 2000 .

[4] Dong-Chul Park,et al. A modular neural network for R2R diagnosis of semiconductor fabrication equipment: a reactive ion etching application , 2004, 2004 IEEE/SEMI Advanced Semiconductor Manufacturing Conference and Workshop (IEEE Cat. No.04CH37530).

[5] B. Flinchbaugh,et al. Monitoring and diagnosis of plasma etch processes , 1988 .

[6] Andrew Skumanich,et al. Automated fault detection and classification of etch systems using modular neural networks , 2004, SPIE Advanced Lithography.

[7] Vladimir N. Vapnik,et al. The Nature of Statistical Learning Theory , 2000, Statistics for Engineering and Information Science.

[8] I. Jolliffe. Principal Component Analysis , 2002 .