水下目标回波盲分离性能的瞬时频率特征评价方法
A performance evaluation methods based on instantaneous frequency feature of underwater target echo blind separation
-
摘要: 在水下主动声呐目标回波与混响盲分离中,针对分离结果顺序不确定性以及缺乏分离有效性衡量手段的问题,提出了以信号瞬时频率特征为指标的盲分离性能评价方法。推导了目标回波与混响的时频分布特性,理论表明目标回波在瞬时频率序列的中心偏离程度以及整体随机程度上低于混响,据此提取信号的瞬时频率方差与瞬时频率熵两种信号特征,并将二者作为从盲分离结果中识别目标回波的依据。海试数据结果表明,在盲分离得到的所有分离信号中,目标回波具有最小的瞬时频率特征值,并且该特征值越小,目标回波与混响的盲分离程度就越高。Abstract: Considering the nondeterministic order of the separation results as well as lack of separation effectiveness measurement approach in blind separation of underwater target echo and reverberation for active sonar, the signal instantaneous frequency feature as the indicator is proposed to be a performance evaluation method. The derivation of the time-frequency distribution character of target echo and reverberation shows that, the center deviation and the randomness of the instantaneous frequency sequence of target echo is lower than those of reverberation. Thus, the signal instantaneous frequency variance and instantaneous frequency entropy as two signal features are extracted to be the evaluation of target echo identification in blind separation results. The sea experimental results have shown that the target echo has the minimum instantaneous frequency feature value among all the separated signals after blind separation, while with smaller feature value, the better separation results of target echo and reverberation will be obtained.