A Hybrid Reliability-Based Design Optimization Approach with Adaptive Chaos Control Using Kriging Model

Abstract

Traditional reliability-based design optimization (RBDO) approaches are computationally expensive for complicated problems. To cope with this challenge, a surrogate-based hybrid RBDO approach for the problems with highly nonlinear constraints and multiple local optima is proposed in this study. Specifically, the adaptive chaos control (ACC) method is used to ensure the robustness of the most probable target point (MPTP) search process, and the particle swarm optimization (PSO) algorithm is utilized to conquer the barrier caused by multiple local optima. Three illustrative benchmark examples together with a 3 m diameter orthogrid stiffened shell for current launch vehicles are employed to demonstrate the efficiency and robustness of the proposed method.