Fuzzy Logic Applications in Diagnosing Mechatronic Systems
Tapio Rauma, Matti Kurki · Studies in fuzziness and soft computing · 1998
Abstract. In this paper, possibilities of using fuzzy system models as a part of model-based fault diagnosis systems are discussed. Methods and experiences of using fuzzy system models at different stages of fault diagnosis systems are presented. In this study, fuzzy system models are used in fault detection and inverse fuzzy system models in fault localization. Inverse fuzzy models are discussed in the light of real-world applications. The mathematical theory is skipped entirely. Additionally we present some examples of using a fuzzy system model in an interface between a fault diagnosis system and a (human) operator. We also briefly discuss a total management of the life-cycle of mechatronic systems. We divide the life-cycle of a mechatronic system into four stages: design, development, installation (integration), and operation and maintenance. When a mechatronic device is being designed and built, it is difficult to analyze the behavior of the device before testing it in practice. It is common that many prototypes of the device are tested at different stages of the development process. The main idea of our approach is that a fuzzy model of a system is built as early as possible, and the model is used, for example, in analyzing the validity of design work done so far. The model is updated and used for different purposes throughout the development and installation process. Finally, the model is used as a part of fault diagnosis and control systems.