声场空频特征非参数融合无人机声学探测

UAV Acoustic Detection Based on Non-Parametric Fusion of Spatial-Frequency Characteristics of Sound Field

针对复杂多源混叠目标声源辨识问题,传统定位算法因存在较多噪声干扰,定位结果会出现很多野点,无法准确地对目标进行估计。本文提出利用高分辨率声成像处理算法获得目标区域声场空频信息矩阵,依据先验目标噪声源频率统计特性得到无人机特征,通过非参数估计的Parzen窗函数法计算空间分布概率密度函数,基于目标在特征频段和空间区域的分布特性,建立空频特征联合优化的检测定位模型。无人机声学检测仿真与实际定位结果表明该方法具有良好的空间抗干扰能力,可实现复杂环境下声源目标的检测定位。

In order to investigate the sound source identification problem with complex multi-sources aliasing, owing to the target signal is mixed with more interference noise, there will be a lot of wild points in the location results of the traditional location algorithm, so it is impossible to estimate the target accurately. the high-resolution acoustic imaging processing algorithm was proposed to obtain the space-frequency information matrix of the sound field in the target region, and the Parzen window function based on non-parameter estimation was designed to calculate the probability density function (PDF) of spatial distribution for wide-band frequency information. After that, a target detection and localization model which jointly optimizes spatial-frequency characteristics is established according to the joint distribution of the target in the characteristic frequency band and the spatial region. The simulation and experimental results show that the proposed method has good anti-interference capability in the UAV acoustic detection application and could realize the effective detection and localization of the sound target under complex environment.