In the present study a methodology to perform fault detection, isolation and estimation is proposed that is based on adaptive model based techniques. Fault detection and diagnostics is performed by comparing the coefficients of healthy system model with that of adapted online coefficients. This approach is shown to be robust to modeling errors, sensor noise and process variability. The proposed approach is applied to FTP-75 cycle simulation data of exhaust gas recircultaion (EGR) faults and is shown to effectively perform fault detection and diagnosis.

This content is only available via PDF.
You do not currently have access to this content.