Model based fault detection of an electro-hydraulic cylinder

One of the key issues in the design of fault detection and diagnosis (FDD) schemes for hydraulic systems is the effect of model uncertainties such as severe parametric uncertainties and unmodeled dynamics on their performance. This paper presents the application of a nonlinear model based adaptive robust observer (ARO) to the fault detection and diagnosis of some common faults that occur in hydraulic systems. The ARO presented in this paper is designed by explicitly taking into account the nonlinear system dynamics. Some robust filter structures are designed to attenuate the effect of model uncertainties and controlled online parameter adaptation helps in reducing the extent of model uncertainty and in increasing the sensitivity of the fault detection scheme to help in the detection of incipient failure. The state and parameter estimates are continuously monitored to detect any off-nominal system behavior even in the presence of model uncertainty. Typical faults in hydraulic cylinders like sensor failure, fluid contamination, and lack of sufficient supply pressure are considered in this paper. Simulation results on the swing-arm of a three degree of freedom hydraulic robot are presented to demonstrate the effectiveness of the proposed scheme.