基于匹配追踪算法的激波信号特征提取

针对时域重叠、噪声混合等情况下从多弹头激波信号中提取时延特征参数问题,利用匹配追踪(Matching pursuits,MP)算法的自适应分解特性对激波混合信号进行特征提取,并针对运行时间长和内存占用大等问题提出"粗提取—精提取"两步改进算法,通过"粗提取"获得各个激波信号的长度,触发位置等先验知识,利用"分库"、"总库"思想重构过完备原子库进行精提取。仿真结果表明,该改进算法不仅能以较高精度提取时延特征参数,而且能有效提高提取速度和减小内存占用。

Feature extraction from the Shockwave signals based on the matching pursuits algorithm

Considering the problem of extracting time-delay feature parameters from multi-warhead shockwave signals in the situation of time-domain overlapping and mixed noise, the adaptive decomposition characteristics of matching pursuits (MP) were used to extract feature from shockwave signals and a two-step improved algorithm used in wide-extraction-precise-extraction was put forward to improve the extraction speed and reduce the memory cost. The length and trigger position of each shockwave was obtained by wide extraction for reconstructing over-complete atom dictionary in next precise extraction based on the part-dictionary and the whole-dictionary. The simulation indicated that the improved algorithm can not only extract time-delay feature parameters at a relatively high precision but also improve the extraction speed and reduce the memory cost effectively.