A New Approach Reliability Based for the Design of Hybrid Expert Systems in Industrial Diagnosis Field

Abstract: Now days, automated factories are ushering a new era of industrial revolution, such factories are becoming more complicated and as a result more difficult for diagnosis operations. The advent of hybrid diagnosis systems in recent years, has allowed the overcoming of some challenging problems related to the application of systems based on a unique model. This paper presents an approach based on reliability methods for the design of hybrid expert systems, which are a new trend in industrial diagnosis systems. First, we introduce some aspects of diagnosis problem complexity in industrial field, and then we present some of the well-known specifications and advantages of hybrid expert systems. Finally, a reliability based approach for the design of such systems is proposed and applied to a milk production unit.