LAUSR

Abstract Various uncertainties, which are usually time-dependent, affect the reliability of complicated engineering systems seriously. Considering the fact that only limited sample data of the uncertain variables can be obtained in engineering practice during the whole in-service time of multidisciplinary systems, the interval process model is introduced to model the time-dependent uncertain variables, and a non-probabilistic time-dependent reliability estimation model is proposed. In addition, a sequential multidisciplinary optimization and non-probabilistic time-dependent reliability assessment (SMO_NTRA) approach is developed to decouple the time-dependent reliability analysis from the multidisciplinary design optimization (MDO). In the framework of SMO_NTRA, the deterministic MDO and non-probabilistic time-dependent reliability analysis are executed in a sequential manner. Thus the computationally expensive double level optimization problem can be avoided and the efficiency can be greatly improved. Furthermore, the shifting distance of the constraint is calculated by bi-section method. Both numerical and engineering examples are employed to demonstrate the validity of the proposed method.