Reliability Index Prediction Using Fuzzy Principle

This paper proposes the method to predict a reliability index of power distribution feeders in the central area of Provincial Electricity Authority (PEA) of Thailand. Using the Weibull distribution, the time to failure data during 2002 2009 is analyzed over moving 4-year periods to determine the failure rates associated with protective devices (Breakers, Reclosers and Fuses). The next year failure rates are then extrapolated using curve fitting. Fuzzy sets are introduced to describe uncertainty of those failure rates to be used in calculation for the future System Average Interruption Frequency Index (SAIFI) using fuzzy arithmetic operation. The resulting SAIFI can be used as a reliability target of the systems for the incoming year.