Abstract: An efficient reliability-based multidisciplinary design optimization(RBMDO) framework for turbine blade was proposed by combining the single-loop-based reliability analysis(SLBRA) method with concurrent subspace optimization(CSSO) method.The Kriging approximation model was used and updated during the optimization process.The proposed method is efficient because it does not need iterative process during finding the most probable points of failure(MPP).A turbine blade design optimization was employed to illustrate the feasibility and efficiency of the proposed method,it shows that the optimal results satisfied the required reliability.From the comparison with the commonly used double-loop based RBMDO method it indicates that the proposed method is much more efficient and well-suited for complex engineering.

Key words: Multidisciplinary design optimization;Concurrent subspace optimization;Single loop method;Reliability analysis;Kriging approximate model;Turbine blade

摘要： 为了得到适用于涡轮叶片复杂结构并同时考虑可靠性的多学科设计优化方法,将基于单循环方法的可靠性分析(SLBRA)与并行子空间设计优化方法 (CSSO)相结合,提出了一种基于可靠性的多学科设计优化(RBMDO)方法。在优化过程中使用Kriging近似模型并不断提高模型精度。该方法在计算最可能失效点(MPP)的过程中避免了优化迭代,提高了计算效率。以涡轮叶片的设计优化为例,对该方法进行了验证并与传统双循环方法进行了对比。结果表明,优化结果满足可靠性的要求,与双循环方法相比优化效率明显提高,证明了该方法在工程应用中的可行性和有效性。

关键词: 多学科设计优化;并行子空间优化方法;单循环方法;可靠性分析;Kriging近似模型;涡轮叶片