In this paper, a data-based covariance benchmark is proposed for control performance monitoring. Within the covariance monitoring scheme, generalized eigenvalue analysis is used to extract the directions with the degraded or improved control performance against the benchmark. It is shown that the generalized eigenvalues and the covariance-based performance index are invariant to scaling of the data. A statistical inference method is further developed for the generalized eigenvalues and the corresponding confidence intervals are derived from asymptotic statistics. This procedure can be used to determine the directions or subspaces with significantly worse or better performance versus the benchmark. The covariance-based performance indices within the isolated worse and better performance subspaces are then derived to assess the performance degradation and improvement. Two simulated examples, a multiloop control and a multivariable MPC system, are provided to illustrate the utility of the proposed approach. Then an industrial wood waste burning power boiler unit is used to demonstrate the effectiveness of the method. © 2007 Elsevier Ltd. All rights reserved.
Bibliographical noteFinancial support for this work from the National Science Foundation under DMI-0432433 and Weyerhaeuser Company through sponsorship of the Texas-Wisconsin Modeling and Control Consortium is gratefully acknowledged.
- Covariance-based performance monitoring
- Data-driven benchmark
- Generalized eigenvalue analysis
- MIMO control performance monitoring
- Performance subspace
- Statistical inference