ON-LINE PROCESS MONITORING USING A ROBUST STATISTICS BASED METHODOLOGY

Abstract

Robust Data Reconciliation strategies provide unbiased variable estimates in the presence of a moderate quantity of measurement gross errors. Systematic errors which persist in time, as biases or drifts, overcome this quantity causing the deterioration of the estimates. This also occurs due to the presence of process leaks. The fast detection of those faults avoids the use of biased solutions of the data reconciliation procedure, and allows to perform quick corrective actions. In this work, a methodology for leak detection is incorporated into a robust data reconciliation procedure that detects and classifies systematic observation errors. The strategy makes use of the Robust Measurement Test, to detect outliers and leaks, and the Robust Linear Regression of the data contained in a moving window to distinguish between biases and drifts. The methodology is applied for two benchmarks extracted from the literature. Results highlight the performance of the proposed strategy.