In this paper, identification and fault diagnosis methods for uncertain Multiple Input Multiple Output (MIMO) Linear Parameters Varying (LPV) models is presented. The fault detection methodology is based on checking if measurements are inside the prediction bounds provided by a MIMO LPV model identified using real data and the parity equations approach. The proposed approach takes into account existing coupling between the different measured outputs. Modeling and prediction uncertainty bounds ar...
In this paper, identification and fault diagnosis methods for uncertain Multiple Input Multiple Output (MIMO) Linear Parameters Varying (LPV) models is presented. The fault detection methodology is based on checking if measurements are inside the prediction bounds provided by a MIMO LPV model identified using real data and the parity equations approach. The proposed approach takes into account existing coupling between the different measured outputs. Modeling and prediction uncertainty bounds are handled using zonotopes. An identification algorithm that provides model parameters and their uncertainty such that all measured data free of faults will be inside the predicted bounds is also proposed. The fault isolation and estimation algorithm is based on the use of residual fault sensitivity. Finally, a case study based on a four tank system is used to illustrate the effectiveness of the proposed approach.
Citation
Blesa, J.; Puig, V.; Saludes, J. Identification and fault diagnosis for LPV uncertain systems. A: IEEE Conference on Decision and Control and European Control Conference. "Proceedings of the 50th IEEE Conference on Decision and Control and European Control Conference". Orlando, FL: IEEE, 2011, p. 3056-3061.
Share
