The TE Fault Monitoring Based on IPCR of Adjustable Threshold.

The The algorithm of Improved Principal Component Regression (IPCR) judges whether there is a quality related fault in Tennessee Eastman (TE) process with \( T^{2} \)-statistics. Because the threshold value is never changed, there will be the problem of false alarm and missing alarm. To solve this problem, an adjustable threshold IPCR algorithm is proposed. Firstly, the IPCR model is built with normal data and the threshold of traditional \( T^{2} \)-statistics is obtained. In the online detection, the new threshold is calculated according to the fixed threshold and the exponentially weighted moving average of statistics, and the new threshold is used for fault detection. Finally, the simulation results in TE process show that this method can effectively enhance the detection results in TE process.