A multi objective evolutionary approach to optimize the design of K-out-of-N systems using AHP technique

In this paper we present an optimization approach to design a configuration including several Kout- of-N systems. In this approach, we employ two incorporated methods. The first method is based on the specification and determination of some criteria and sub criteria to rank systems based on their performance during operational time according to OREDA handbook. The high ranked systems are considered as primary selected configurations among many possible and practical systems. The second one is a multi objective modeling with determination of effective constraints such as maximum cost limitation. Because of nondeterministic data, we propose failure rate of each safety instrumented system as random variables and use their expected values as input parameters of the model. These values and the failure rates' variances are calculated with new acceptable method. The result of this method and the primarily ranking of system safety configurations from AHP technique synthesize with multi objective evolutionary approach to present a new solution for K-out-of-N system design problems.

A multi objective evolutionary approach to optimize the design of K-out-of-N systems using AHP technique Keywords:

K-out-of-N systems , Redundancy allocation , Analytic hierarchy process , Multiobjective optimization