FMEA of Turbocharged Diesel Engine System Using Fuzzy Inferencing 2000-01-0521
This paper presents a novel method of the FMEA of the turbocharged diesel engine system using fuzzy inferencing. It uses a membership function to represent information of FMEA (e.g. failure mode and failure effect) that is described by linguistic variables to the defined categories. The knowledge of experts is expressed in terms of if-then rules. Min-max inferencing is used to evaluate fuzzy model, and the Center of Area (centroid) method of defuzzification is adopted to obtain crisp results.
A fuzzy-logic-based FMEA Two-stage Inferencing Model is developed. A prototype system based on the model has been constructed and used to conduct the FMEA of the turbocharger rotor-bearing subsystem. The results have shown the practicality of the method.