空中目标信号相空间重构方法研究

为提取空中目标辐射噪声非线性和混沌特征,需要对目标信号作非线性时间序列分析,其第一步则是相空间的重构,基于Takens定理进行相空间重构前必须先确定重构相空间的时滞和嵌入维数两个重要参数。分别采用互信息方法确定重构最佳时延和关联积分法确定重构嵌入维数。分别以经典混沌信号和实测目标噪声数据为研究对象进行计算仿真,结果证明了该方法选择地参数进行相空间重构的有效性与准确性,重构的相空间能很好反应原混沌系统的特性,为下一步分析提取目标非线性特征奠定了良好的基础。

Study on Parameters Selection for Phase Space Reconstruction of Aerial Target

For extracting the nonlinear and chaotic features of aerial target, nonlinear time series analysis was considered for the target signal, and the first step was phase space reconstruction. Selection of the optimal delay time and embedding dimension parameters was the key based on Takens theorem. Mutual information method and correlation integral method were presented to get the two optimal parameters. Take typical chaotic signal and measured target noise data as study objects, simulation results show that the reconstructed phase space is accurate and effective based on the selected parameters from this method. It can response characteristics of the original chaotic system effectively, and makes a good preparation for the next study of nonlinear features extraction for target signal.