中文摘要：

基于瞬时特征提取的数字信号自动识别方法具有重要的应用前景。采用此方法识别信号时，需要提取信号的瞬时幅度、瞬时相位、瞬时频率去计算分类特征参数。但提取信号瞬时相位时会出现相位卷叠现象。本文首先对相位卷叠和相位去卷叠进行了详细的分析，然后提出了一种估计相位去卷叠理论门限范围的方法。而其它相关文献都没对这个门限进行过深入的研究。MATLAB仿真结果表明，采用本文估计出的去卷叠门限范围，可以直接有效地提取信号的非线性瞬时相位。

英文摘要：

Digital signal automatic identification based on extracted instantaneous feature has important applications. While adopting the method to identify signal, we must extract the instantaneous amplitude, phase and frequency of the signal to calculate the classification parameters. But phase wrapping occurs when extracting instantaneous phase. This paper analyzes phase wrapping and unwrapping in detail,then proposes a method to estimate the theoretical threshold range of phase unwrapping. And other relevant references didn＇ t carry out an in-depth study on the threshold. Simulation results of matlab show that non-linear instantaneous phase can be effectively and directly extracted if adopting the phase unwrapping threshold range estimated in this paper.