Diagnostic by luenberger observer using bond graph approach for a photovoltaic system

Abstract

This article deals with the design of a new methodology for diagnosis based on bond graph modeling. The observer design provides an estimated state from the model and inputs and outputs of measurements. The luenberger observer diagnosis is determined by exploiting the leap graph architecture to generate the residuals. In addition, the performance of the diagnostic system with the proposed luenberger observers is used to analyze the robustness of the system against the defects served. The results of the research were applied to a photovoltaic system. In this context, the proposed method was carried out from the modeling step of the system to the diagnosis step. The paper concerns the design of a new methodology for diagnosis based on bond graph modeling. The observer design provides a state estimated from the model and inputs and outputs measurements. Furthermore, we have exploited the architectural of the bond graph to generate the diagnostic condition based on Luenberger observers.