Reliable system design under uncertainty

Abstract

System design often concerns performance while ignoring reliability, leading to a satisfactory but unreliable outcomes. To overcome this problem, system design combining reliability allocation is proposed, and implementing under uncertainty. Analytical target cascading (ATC) is used as the basic framework to conduct the procedure using a sequential optimization and reliability assessment (SORA) method. ATC could decompose the system into a multilevel structure corresponding to its reliability model, and SORA improves the efficiency of the whole design optimization. The gist of the proposed method is to involve reliability into variable set and operate optimization in such ATC multilevel system under probabilistic uncertainty, resulting in coordinated system design and reliability allocation. By this mechanism, reliability affects design to obtain a compromise. At last, a structural missile problem is tested to validate the effectiveness of the proposed method.