通过分析基于响应面的并行子空间优化算法的特点指出并行子空间优化算法学科级优化的作用在于向系统级优化响应面提供性能优良的设计点.在此基础上,建立了不要求学科级优化的改进的并行子空间优化算法,进一步降低了设计优化的计算量,解决了分析模块与优化模块间的接口困难.依据该算法建立了结构、气动和隐身一体化设计优化框架,实现了某无人机机翼计及气动和隐身约束的结构综合优化.
After the analysis of Concurrent Subspace Optimization (CSSO) process, this paper points out that the function of disciplinary optimization of CSSO is to offer experiment points of response surface of system optimization. Based on this conclusion, an Updated Concurrent Subspace Optimization (UCSSO) which does not require disciplinary optimization but uses Uniform Experiment Design to arrange experiment points is presented. According to the updated algorithm, the framework of structural optimization with constraints of aerodynamics and RCS of aircraft is researched. The result of structural optimization with constraints of aerodynamics and RCS to an unmanned aircraft wing proves that UCSSO can greatly reduce calculation time and decrease the difficulty of data exchange between discipline.