Sensor auditing

This chapter introduces two sensor audit strategies that can detect and diagnose sensor faults. The first strategy (Section 8.1) focuses on sensor auditing by using calibration and test data sets that are processed either by developing PLS models (for data with low autocorrelation) or canoni-... 

CVSS models introduced in Section 4.5 for process data with strong autocorrelation and crosscorrelation. One-step ahead residuals generated from the CVSS model will be used for sensor audit. A multipass version of the technique is developed to identify submodels by eliminating successively the corrupted measurements from both the calibration and test data sets and fitting a different CV state-space model. 

Figure 8.1. The multi-pass PLS or CVSS sensor audit algorithm for detecting simultaneous faults suggested in [290].

A bias is added to the reactor conversion measurement with a magnitude of about 5 % of the current reactor conversion (sensor 2). Immediately after the bias change is introduced to the sensor, both the univariate chart for reactor conversion and the multivariate and SPE charts (Figure 8.7) indicate an abnormality. A CVSS model is developed to generate the residuals for sensor audit. The KBS automatically begins the sensor validation routine to determine the source cause of the inflated and SPE statistics. Figure 8.8 shows that the residuals mean for sensors 1 (initiator concentration), 2 (reactor conversion), and 4 (polydispersity) have exceeded... 

A Negiz and A Cinar. Monitoring of multivariable dynamic processes and sensor auditing. J. Process Control, 8(5-6) 375-380, 1998. 

E Tatara. An Integrated Knowledge-Based System for Automated System Identification, Monitoring, and Sensor Audit for Multivariate Processes. Master s thesis, Illinois Institute of Technology, Chicago, IL, 1999. 

Parameter change detection, 27 Partial least squares, 42, 79 convergence, 81 inner relations, 81 multi-block, 113 multipass PLS for sensor auditing, 204... 

---