目的　研究用于识别直升机目标声信号的特征提取方法 .方法　通过对直升机声信号频域特性分析 ,采用基于子空间分解的多重信号分类法 ( MUSIC)算法提取信号谐波频率作为目标特征 ,利用子空间分解将观测数据分解为信号子空间与噪声子空间的特点 ,抑制噪声干扰 ,提高识别能力 .结果与结论　对实测的直升机声信号谐波频率提取结果表明 ,子空间分解方法提取目标频域特征是可行的 ,用 MUSIC算法提取谐波频率值是有效的 Aim To study the feature extraction method of helicopter target acoustic signal.Methods Based on the analysis of frequency characteristics of helicopter acoustic signal, the harmonic signal was extracted by MUSIC algorithm based on subspace decomposition as the target feature, Spatial analysis decomposes the observed data into the characteristics of signal subspace and noise subspace, suppresses the noise interference and improves the recognition ability.Results and Conclusions The experimental results of the harmonic frequency of helicopter acoustic signals show that the subspace decomposition method extracts the target frequency domain features It is feasible to use the MUSIC algorithm to extract the harmonic frequency value is valid