Data Driven Evaluation of Nonlinear Output Frequency Response Functions with Applications to Structural System Fault Diagnosis

In this study, the Nonlinear Output Frequency Response Functions (NOFRFs) is evaluated from a data driven models of nonlinear systems. For this purpose, a Nonlinear Auto-Regressive with eXogenous inputs (NARX) model with input nonlinear terms is identified so that the system NOFRFs can be evaluated up to an arbitrarily high order without involving issues with truncation errors. Moreover, the application of the data driven model based NOFRFs evaluated from the data driven NARX model can well extract fault related system characteristics and facilitate structural system fault diagnosis.