An improved thresholding technique for fault detection in a cement rotary Kiln

Abstract

On-line control of nonlinear nonstationarity processes using multivariate statistical methods has recently prompt a lot of interest due to its industrial practical importance. Indeed, basic process control methods do not allow monitoring of such processes. For this purpose, this study proposes a moving window real-time monitoring system based on a Principal Component Analysis scheme. A sliding window of a fixed length is used to determine an appropriate threshold. This thresholding technique will provide an accurate detection. At each one incoming observation, the developed on-line threshold will avoid false alarms and increase non detection faults. It will be shown through experimental results that