地面装甲目标声信号的混沌特征提取

针对地面装甲目标辐射的噪声信号的非线性特性，为使智能地雷能够有效地识别目标，利用非线性动力学理论中的混沌原理对目标声信号进行特征提取。通过野外场地实验，采集到2种装甲目标在不同运行速度下的40组样本信号，采用改进C-C法求得信号时间序列的相空间重构参数——时延和嵌入维，再利用Wolf法得到了2种目标声信号的混沌特征量——最大Lyapunov指数。结果显示：同一目标声信号的最大Lyapunov指数相近，且与运动状态相关性不大；不同目标间声信号的最大Lyapunov指数相差较大，辨识度较高。结论证明，最大Lyapunov指数可以作为地面装甲目标识别的有效特征参量。

Chaotic feature extraction of acoustic signals from armored vehicles

The character of acoustic signal radiated from armored vehicles are proved to be nonlinear．To effectively identify the armored vehicles，chaos theory based on nonlinear dynamics was used for extracting the feature of acoustic signals．40 sample signals of two kinds of armored vehicles running in different speeds were collected by outdoor experiment．The reconstruction parameters (time delay and embedding dimension) of phase space from the time series were obtained by the improved C C method，then the largest Lyapunov exponents for each signal were calculated with these parameters by Wolf method．The results show that the values of the largest Lyapunov exponents from the same target are so close，while varying significantly between two different targets，and that the speed of the targets has no obvious impact on the largest Lyapunov exponents．It indicates that the largest Lyapunov exponent can be the characteristic parameter in target identification of acoustic signals for smart landmines．