Abductive Approach Using Fuzzy Petri Nets for Industrial System Diagnosis

In an industrial decentralization context of control and migration of intelligence downwards, mechanical equipments with more and more electronics and automations have been developed given the so called mechatronic systems. However this rapid progression of new technology has contributed to increase the systems complexity and decrease the mastering. The industrial diagnosis associated at a precocious detection of degradations have a key role in quality mastering, availability improve and productivity of the production tool. The industrial diagnosis methods are divided in two categories: diagnosis methods with formal model of the equipment and diagnosis methods without equipment model. The formal mathematical models of industrial equipments are often with uncertainties and difficult to built. This paper proposes a method based explanatory model which is best suited to model the cause-effect relationships essential for the industrial diagnosis. We adapt a novel tool of assistance to diagnosis based on the fuzzy Petri nets (FPN). An industrial application concludes this work.