USING ENVIRONMENTAL MODELS APPROXIMATED BY FUZZY IDENTIFICATION FOR HYBRID PLANNING OF MECHATRONIC SYSTEMS

摘要

Above the controller level a lot of components are needed in mechatronic systems for the development towards self-optimizing systems. One of these components is a hybrid planning architecture. This architecture integrating discrete and continuous domains is of major importance to support the permanent determination of system objectives and their implementation during the course of action. Through this the principle of self-optimizing mechatronic systems is defined as well. Such a novel hybrid planning architecture is outlined in this paper. In order to plan efficiently and safely, environment models are needed for predicting future system behaviors. In this paper we propose a fuzzy logic based approach to environment modeling and apply it in a railway-bound domain within the context of an air gap adjustment system for a dual-fed linear motor powering a wheeled train.