摘要
为了改善协同优化方法的收敛性能并提高求解效率,提出了一种基于径向基函数代理模型的改进协同优化方法。该方法采用基于光滑参数优化径向基函数代理模型,构造了协同优化方法系统级约束的近似模型,并将置信域和均匀设计方法相结合,完成近似模型的不断更新,引入粒子群优化算法完成了系统级和学科级优化问题的求解。利用数值计算和减速器设计两个典型算例对所提方法进行了验证,求解结果表明了改进协同优化方法可行且有效。
In order to improve convergence performance and computational efficiency of the conventional collaborative optimization method, an improved collaborative optimization method based on radial basis function surrogate model is proposed in this paper. In this method, the approximate optimization models of constraint conditions in system-level are constructed using radial basis function surrogate model, and are updated by uniform design method combined with confidence regions, and the particle swarm optimi- zation algorithm is introduced to the system-level optimization and disciplinary-level optimization. The im- proved collaborative optimization method is tested by using numerical calculation and decelerator design as two typical examples. The process and result verify the feasibility and effectiveness of the proposed method.
出处
《西安理工大学学报》
CAS
北大核心
2013年第3期273-278,共6页
Journal of Xi'an University of Technology
基金
国家自然科学基金资助项目(60903124)
西安理工大学科学研究基金资助项目(102-211112)
关键词
改进协同优化方法
径向基函数
代理模型
粒子群优化算法
improved collaborative optimization method
radial basis function
surrogate model
particleswarm optimization algorithm
