Reliability allocation weights using ME- OWA method with genetic algorithm

This paper presents two approaches to reliability allocation and makes an attempt to improve the approximation of reliability allocations that was obtained in previous studies. Reliability Allocation goals to specify the setting of reliability aims for individual subsystems to meet the specified reliability goal and to balance the hardware and software subsystem goals. Well-balanced sometimes refers to approximate relative equality of development time, difficulty, risk, or to the decrease of overall development cost. The apportionment of reliability values between the various subsystems and elements can be generally made on the basis of complexity, criticality, estimated achievable reliability, or any other factors considered appropriate by the analyst. System-level allocations are successively decomposed applying the reliability model(s) until an appropriate set of reliability measures has been allocated to each hardware and hardware/software component of the system. In first section, ME-OWA and FOO methods are briefly introduced. In second section, relevant equations of ME-OWA are solves and discussed using genetic algorithm (GA) and as a result the best approximation of reliability allocations are reported. Finally, a practical example is presented. The results show that the proposed method can accurately and efficiently allocate reliability ratings throughout reasonably assigned reliability levels in subsystems, meet customer needs, control reasonable support costs, and decrease maintenance costs.