基于Kriging代理模型的倒角建筑流场优化

Optimization of Chamfer Angel Architecture Flow Distribution Based on Kriging Surrogate Model

结构高层建筑受风荷载的影响较大,而建筑形态是影响建筑风荷载的一个重要因素。故建立了倒角建筑流场优化的4种代理模型,通过误差分析发现Kriging代理模型是较好的代理模型,而后采用MIGA,NLPQL,AMGA,MMFD的4种算法对代理模型进行了优化分析,使模型中的最小负压最少提高了48.09%,达到了优化目标。

Structural high- rise buildings have been affected by the wind load, which has been influenced by architectural form. This passage built four surrogate models optimized by chamfering architecture flow field. By the error analysis, we found that Kriging surrogate mode is better than others, and later adopted four algorithms, MIGA, NLPQL, AMGA and MMFD, for surrogate optimized analysis, which has increased by 48.09% at least of the model minimum negative pressure to optimize objective.