Design Optimization under Aleatory and Epistemic Uncertainties

Deterministic design optimization (DO) may lead to unreliable designs due to not taking into consideration uncertainties. In recent years, DO under uncertainty has attracted more and more attention. In DO practice, both aleatory and epistemic uncertainties may exist, furthermore, the two types of uncertainties may affect both the objective function and the constrains of the optimization problem. In this paper, we develop a DO model which could cater for the above-mentioned situation. We also propose a new algorithm, sequential optimization and reliability and possibility assessment (SORPA), to solve the developed DO model. Numerical example is given, and the results obtained demonstrate the feasibility of the developed DO model and the efficiency of the proposed algorithm.