Identification et diagnostic des systèmes linéaires à paramètres variants (LPV) application a un système thermique

Abstract

In this work, we demonstrated the usefulness of Linear Parameter Varying representations (LPV) and their ability to approximate the Nonlinear Systems. For the identification part, the problem of deriving a dynamic model for a MIMO actual heat exchanger is considered. To take into account the dynamic system dependency on the mass flow rate of hot and cold side explicitly, a continuous time input-output LPV model is used. In this case, the parameters of the LPV model are identified by a local approach using specific methods of nonlinear programming to minimize the error functions. Simulations are performed and the results clearly demonstrate the advantages of the proposed approach. For the fault diagnosis part, the diagnostic system has been constructed for the detection and estimation of faults in the different parts of the monitored system. Various new techniques have been used for detecting, isolating and estimating of faults. The simulation results clearly show the effectiveness of the proposed methodology.