计算机仿真已成为现代工程设计的重要手段，随着分析模型越来越精细，各种不确定因素在设计中也被加以量化。为了得到可靠且性能更加优异的设计方案，基于可靠性的优化（Reliability-Based Optimization, RBO）已成为现代设计的发展趋势。然而，在面对耗时仿真模型时，RBO中相互耦合的可靠性分析和优化过程使得问题求解面临巨大的挑战。本项目以飞行器设计中有限元耗时仿真模型RBO为背景，基于协同径向基函数方法，分别从可靠性分析、优化及其相互耦合策略入手，通过特定的准则识别包含最优点且易失效的局部重要区域，构造高精度的局部代理模型，将原问题转化为基于Co-RBF可靠性分析和约束平移后的确定性优化问题，从而减少耗时模型的总评估次数。本项目最终成果将RBO计算耗时和精度控制在可接受的范围，有望解决耗时仿真模型RBO中大规模计算瓶颈问题，对耗时仿真模型的现代可靠性设计产生积极推动作用。

Numerical simulation has been an important technique in modern engineering design. With higher and higher accuracy, different uncertainties are considered and quantified in the design. To obtain a more secure and reliable design scheme with high performance, designers gradually start to focus on Reliability-Based Optimization (RBO). However, due to the complex and time-consuming model analysis, reliability analysis process and optimization process nesting in RBO face a huge challenge. This program comes from the background of RBO of aircraft design with aerodynamic and structure simulations. Cooperative Radial Basis Function (Co-RBF) is adopted in different processes of RBO, including reliability analysis, optimization and the coupling strategy. Special criterion are adopted to locate the important region including optimal point close to failure boundaries, such that the nested RBO problem is transformed into decoupled determined optimization and reliability analysis problems with surrogate model. Therefore, the total number of expensive simulation estimations reduces a lot. The final progeny of this program is expected to reduce the computation to an acceptable range with reliable accuracy, which facilitates the development of modern reliability-based design with time-consuming simulation model.