基于代理模型的麦克风相位阵列设计技术研究  

Research on microphone phase array design based on surrogate model

作　　者：丁存伟 李周复[1] 张雪[1] 周国成[1] Ding Cunwei;Li Zhoufu;Zhang Xue;Zhou Guocheng(Aerodynamics Research Institute,Aviation Industry Corporation of China,Harbin 150001,China)

机构地区：[1]中国航空工业空气动力研究院,哈尔滨150001

出　　处：《实验流体力学》2018年第4期93-98,103,共7页Journal of Experiments in Fluid Mechanics

摘　　要：气动声学风洞试验过程中,针对目标声源特性进行麦克风相位阵列改进时,为保证风洞试验效率,必须在较短的时间内完成阵列优化设计工作。为满足这一试验需求,最大限度提升麦克风相位阵列设计效率,引入了基于Kriging代理模型的优化设计方法。采用基于点扩散函数的计算程序进行阵列性能分析,获取阵列最大旁瓣水平和分辨率。通过对样本点响应值进行计算,建立Kriging代理模型,进而以计算速度极高的Kriging代理模型作为阵列性能分析方法开展优化搜索,避免了大量调用阵列性能计算程序导致计算耗时过高的问题,显著提升了麦克风相位阵列设计效率。该阵列设计方法能够有效改善麦克风阵列的测量性能,满足声学风洞试验的特殊应用需求。In aero-acoustic wind tunnel tests,the improvement of the microphone phase array according to the noise source features must be conducted as fast as possible to ensure the required efficiency of the tests.An optimization method based on the Kriging surrogate model is proposed in this study.The maximum sidelobe level and the resolution of the microphone phase array are obtained by analyzing the results computed from the point spread function.A Kringing surrogate model is constructed based on the response values computed on sample points.Then the surrogate model was utilized to fast assess the maximum sidelobe level and the resolution of the microphone phase array,which avoided the high computational cost caused by computing the point spread function.This method can improve the performance of the microphone phase array effectively for aero-acoustic wind tunnel tests.

关 键 词：麦克风相位阵列 优化方法 代理模型 分辨率 最大旁瓣水平 

分 类 号：V211.7[航空宇航科学与技术—航空宇航推进理论与工程]