基于瞬时频率的 Duffing 振子相变判别新方法

A New Method for Phase Transition Identification of Duffing Oscillator Based on Instantaneous Frequency

利用Duffing 振子检测微弱信号的关键在于对振子相变的正确判别,针对现有Duffing 振子相变判别方法存在计算量大、不易量化处理的问题,本文利用经验模态分解方法,研究了混沌临界状态以及大尺度周期状态下Duffing 振子系统输出信号瞬时频率特性,提出了以系统输出第一层内蕴模态函数瞬时频率极差为度量的相变判别方法。基于此相变判别方法,提出了一种新的基于瞬时频率的Duffing 振子微弱信号检测方法。通过仿真计算,实现了信噪比(SNR)为-52dB 的微弱正弦信号的Duffing振子检测。

Phase transition identification is the key technology of weak signals detection using Duffing oscillator. Considering the complexity and huge computation of existing algorithms, the instantaneous frequencies of the chaotic and large scale period states of Duffing oscillator system output were analyzed using empirical mode decomposition. A novel method for phase transition identification based on instantaneous frequency of the first intrinsic mode function of Duffing oscillator system output was proposed. Based this phase transition identification algorithm, a new method of weak signals detection using Duffing oscillator was proposed. Simulation results indicate this method’s validity.