[Objectives] It is time-consuming to solve the robust optimization problem with interval uncertainty if the nested differential evolution algorithm is directly used. A new robust design optimization method is proposed.
[Methods] In the proposed method, the individuals’ response values that have been accurately calculated within the critical distance are used to approximately predict the response values of other individuals and to evaluate the robustness indexes accordingly. Besides, the information of the accurate individuals’ response values, which is gradually expanded in the evolutionary procedure, is used to selectively re-evaluate the robustness of some past individuals, and the critical distance is adaptively reduced based on the misjudgment rate of the robustness.
[Results] Two numerical and one engineering examples are tested to demonstrate the applicability of the proposed algorithm. The results show that the proposed algorithm saves more than 94% of the computational resources while the estimated error is less than 2.5%.
[Conclusions] The proposed method can greatly reduce the times of individuals’ response values calculation in the process of evolution, and keep the adaptive balance between the accuracy and cost for robustness evaluation by using previous optimization knowledge, which provides a new idea and method for robust design optimization with interval uncertainty.