电子测量与仪器学报

赵慎,诸皓冉,周超,李伟,张锐.声学探测无人机中的麦克风立体阵列优化设计[J].电子测量与仪器学报,2025,39(5):155-165

声学探测无人机中的麦克风立体阵列优化设计

Optimization design of microphone array for acoustic detection drones

DOI：

英文关键词:array configuration acoustic detection multi-objective optimization selection strategy peak side-lobe level

基金项目:国家自然科学基金青年基金项目（62103426）、湖南省教育厅科学研究重点项目（23A0464）、湖南省研究生科研创新项目（QL20230271）、湘江实验室重大项目（23XJ01003）资助

作者	单位
赵慎	湖南工商大学智能工程与智能制造学院长沙410205
诸皓冉	湖南工商大学智能工程与智能制造学院长沙410205
周超	国防科技大学智能科学学院长沙410073
李伟	湖南工商大学智能工程与智能制造学院长沙410205
张锐	湖南工商大学智能工程与智能制造学院长沙410205

Author	Institution
Zhao Shen	School of Intelligent Engineering and Intelligent Manufacturing, Hunan University of Technology and Business, Changsha 410205, China
Zhu Haoran	School of Intelligent Engineering and Intelligent Manufacturing, Hunan University of Technology and Business, Changsha 410205, China
Zhou Chao	School of Intelligent Science and Technology, National University of Defense Technology, Changsha 410073, China
Li Wei	School of Intelligent Engineering and Intelligent Manufacturing, Hunan University of Technology and Business, Changsha 410205, China
Zhang Rui	School of Intelligent Engineering and Intelligent Manufacturing, Hunan University of Technology and Business, Changsha 410205, China

摘要点击次数: 1829

全文下载次数: 914

中文摘要:

在基于声学信号探测与定位低空无人机的应用中，针对现有规则平面麦克风阵列测向空间分辨力低、抗干扰能力差等问题，建立适用于对无人机测向的随机立体阵列多约束优化设计模型，并提出基于精英锦标赛筛选策略的优化求解方法。以四环立体阵列为基础阵列构型，最小化峰值旁瓣电平为优化目标，设定阵列结构约束条件，同时限定波束主瓣宽度，构建多参数约束的随机立体阵列优化模型。进一步，将精英策略与锦标赛策略相结合，提出一种多融合筛选策略，应用于遗传算法迭代求解过程中，提升算法的收敛性和全局搜索能力，以得到优化设计后的目标阵列构型。仿真和实验表明，目标阵列的探测方向图中“野点”数量较少，表现出较好的抗噪声性能和空间分辨力，与四环立体阵列相比，目标阵列对低空无人机的探测失效率降低了4.33%，方位角误差与俯仰角误差分别降低了1.54°与0.73°，最远探测距离提升了12 m。

英文摘要:

In the application of detecting and locating low-altitude drones using acoustic signals, the existing planar microphone arrays face problems such as low directional resolution and poor interference resistance. The research focuses on addressing these challenges by establishing a stochastic three-dimensional(3D) array optimization design model with multiple constraints, suitable for drone direction finding. Additionally, an optimization-solving method based on an elite-tournament selection strategy is proposed. Based on a four-ring 3D array configuration, the model minimizes the peak side-lobe level as the optimization objective while setting array structural constraints and limiting the beam main lobe width. A multi-parameter constraint optimization model for stochastic 3D arrays is constructed. Furthermore, an elite-tournament selection strategy is proposed for optimizing the solution process. The elite strategy and tournament strategy are combined into a multi-fusion selection strategy, which is applied during the iterative process of genetic algorithm optimization. This combination enhances the convergence rate of the algorithm and its global search capability, leading to the achievement of the optimized array configuration. Simulation and experimental results show that the direction finding pattern of the target array exhibits fewer false detection points, demonstrating improved noise immunity and spatial resolution. Compared to the four-ring 3D array, the target array reduces the detection failure rate for low-altitude drones by 4.33%, and the azimuth and elevation angle errors decrease by 1.54° and 0.73°, respectively. The maximum detection distance is improved by 12 m.

查看全文查看/发表评论下载PDF阅读器