摘要
针对新型作战体系下以装甲车辆为主的地面目标的被动声识别问题,为实现不同车型在不同工况下的声识别,以常见的3种坦克和4种履带式装甲车为识别对象,提出了一种基于变分模态分解(Variational Mode Decomposition,VMD)和人工蜂群(Artificial Bee Colony,ABC)算法优化的支持向量机(Support Vector Machine,SVM)相结合的装甲车辆声识别模型。首先,采集不同工况下的车辆噪声信号并进行频谱分析,证明了VMD分解的可行性;其次,对样本信号进行VMD分解,得到不同尺度的本征模态函数(Intrinsic Mode Function,IMF)并进行多尺度模糊熵(Multi-scale Fuzzy Entropy,MFE)的计算,得到多尺度模糊熵特征(VMD-MFE);然后,利用优化算法对SVM进行优化,得到最优参数优化的分类器模型;最后,对噪声信号进行特征提取和分类实验。结果表明:VMD的分解效果优于经验模态分解(Empirical Made Decomposition,EMD)和集合经验模态分解(Ensemble Empirical Mode Decomposition,EEMD);与引力搜索算法(Gravitational Search Algorithm,GSA)和布谷鸟搜索(Cuckoo Search,CS)算法相比,ABC算法得到的优化模型ABC-SVM具有更高的识别率,可达94. 14%以上。
For the problem of passive acoustic recognition of ground targets based on armored vehicles under the new combat system,in order to realize the acoustic recognition of different models under different working conditions,taking the common three kinds of tanks and four types of tracked armored vehicles as recognition object,a kind of acoustic recognition model of armored vehicle is proposed based on the Variational Mode Decomposition( VMD) and the Support Vector Machine( SVM) optimized by the Artificial Bee Colony( ABC) algorithm. Firstly,the vehicle noise signal under different working conditions is collected and analyzed by the spectrum,which proves the feasibility of VMD decomposition. Secondly,the sample signal is decomposed by the VMD to obtain the Intrinsic Mode Function( IMF) in different scales,and the Multi-scale Fuzzy Entropy( MFE) is calculated to obtain the VMD-MFE features. Then,SVM is optimized by the optimization algorithm to obtain the classifier model of optimal parameter optimization. Finally,feature extraction and classification experiments are carried out for noise signal. The results show that the decomposition effect of VMD is better than EMD and EEMD. Compared with Gravity Search Algorithm( GSA) and Cuckoo Search( CS),the optimization model ABC-SVM obtained by ABC Algorithm has a higher recognition rate,reaching more than 94. 14%.
作者
樊新海
石文雷
张传清
FAN Xin-hai;SHI Wen-lei;ZHANG Chuan-qing(Vehicle Engineering Department,Army Academy of Armored Forces,Beijing 100072,China)
出处
《装甲兵工程学院学报》
2018年第6期68-73,共6页
Journal of Academy of Armored Force Engineering
基金
军队科研计划项目
