Abstract:

This paper deals with the application of model-based fault detection in diesel-hydraulic drives using the observer-based approach. Based on the nonlinear adaptive observer theory, residual generators are developed which take into account system nonlinearity, model uncertainty and disturbances. This adaptive observer-based FDI scheme possses robustness to the model uncertainty and disturbances and provides a new way to deal with FDI in nonlinear systems. Simulation studies demonstrate that faults in the diesel-hydraulic drives can be successfully detected with this technique.