Reliability-based multidisciplinary concurrent design optimization method for complex engineering systems

Abstract

Reliability-based multidisciplinary design optimization (RBMDO) has been considered as an efficient technology for addressing complex engineering design optimization problems under uncertainties. Current RBMDO methods, however, perform design optimization either in a sequential manner or in an all-at-once manner, which does not support design autonomy in various disciplines and has low computational efficiency. In this article, a reliability-based multidisciplinary concurrent design optimization (RBMCDO) method is proposed to solve the problems of lack of design autonomy and low computational efficiency in current RBMDO methods. In RBMCDO, the multidisciplinary reliability constraints are first transformed into deterministic constraints using the multidisciplinary probabilistic constraint transformation method, developed in this work based on the performance measure approach and Karush–Kuhn–Tucker conditions. The RBMDO problem is then decomposed into several optimization problems of a single discipline, which are solved in parallel based on the collaborative optimization strategy. Two examples are used to verify the proposed method.