Fault diagnosis and prognosis of a hydro-motor drive system using priority valve

Abstract

In this article, an improved fault isolation in the model-based fault detection and isolation (FDI) method is presented using parallely computed bond graph models. All component faults in a system may not be uniquely isolable. However, some faulty parameters’ subspaces may be identified. One of the possible solutions as proposed in this article is to estimate parameters from the actual performance of the plant assuming a single fault hypothesis. Then incorporating the estimated values, all the parallel models are run on a single-core processor and their responses are compared with the healthy plant to identify the actual fault. In this respect, a typical hydro-motor drive system is considered for the FDI analysis, where a stable source of flow is supplied through a priority flow divider valve to two hydro-motors. A methodology for parametric fault isolation under single fault assumption with minimum measurements is discussed. Such method is capable to estimate the remaining useful life (RUL) of the faulty components of a system. The proposed methodology used for identifying the fault and estimating the RUL is simple enough to adopt it in industrial practices.