Functional Model And Its Reliability Analysis Applied Research

Functional models represent not only structural and behavioral knowledge but also functional and teleological knowledge. Therefore, the ability of functional models to represent knowledge is beyond the ability of traditional models. During the past decades, great development is taking place on functional models. They have been widely used in many industrial domains, especially fault diagnosis domain.Three functional modeling strategies named GTST, MFM and Fdef that are widely used in industry are discussed in this dissertation. GTST and Fdef are functional modeling methods based on functional primitives. However, MFM describes the system in term of flows of mass, energy and information. GTST and MFM are often used as a basis of fault diagnosis on nowadays and even in the future, while Fdef is no longer used so often because it has some unavoidable drawbacks.As a result of thorough study on different modeling approaches, the dissertation proposes a new functional modeling method for fault simulation and fault diagnosis. It starts with the inspiration of object-oriented technology and can model the system in the bottom-upper order. The functional model is hierarchical and modular. It can describe the structural, behavioral and functional knowledge of system in a single model. Moreover, the relation between behavior and function is also created in the model.Finally, a functional model of the steering system of some type of tank is established according to the new method. A prototype named FailSim of fault simulation is developed. The result of simulation is used to show validity of the modeling method and therefore to be the basis of following work, such as fault diagnosis.