论文部分内容阅读
为了更好地解决飞行器多学科设计优化问题,对传统基于响应面的并行子空间优化算法(RS-CSSO)进行改进:改进基于近似模型,在具有同等计算精度的情况下减少学科分析的次数,采用均匀试验设计代替学科级优化来直接获得性能优良的初始设计样本点;在系统级优化过程中引入自适应近似模型算法,在迭代过程中对两种近似模型的精度进行对比,以相对误差的均值和标准差作为判据,选用精度更高的近似模型来提高系统级优化效率。采用改进的RS-CSSO算法对飞翼布局无人机进行了设计优化,并与传统RS-CSSO算法进行了对比。结果表明,改进的RS-CSSO不但有着更小的计算量,而且得到了更优的结果,可以应用于飞行器设计多学科优化。
In order to solve the multi-disciplinary design optimization problem of aircraft better, the traditional response surface-based parallel subspace optimization algorithm (RS-CSSO) is improved. Based on the approximate model, the number of disciplinary analysis with the same precision is reduced, The uniform design of experiment is adopted instead of subject-level optimization to obtain the initial design sample points with good performance directly. In the process of system-level optimization, an adaptive approximation model algorithm is introduced to compare the accuracy of two approximate models in iterative process, Average and standard deviation as a criterion, the selection of higher accuracy approximation model to improve system-level optimization efficiency. The improved RS-CSSO algorithm is used to design and optimize the flying wing layout UAV and compared with the traditional RS-CSSO algorithm. The results show that the improved RS-CSSO not only has a smaller amount of computation, but also obtains better results, which can be applied to multidisciplinary optimization of aircraft design.