基于信息熵的自适应窗长STFT算法在AFVISAR中的应用

摘要

Instantaneous frequency estimation was performed on the output signal of all-fiber velocity interferometer for any reflector (AFVISAR) with short-time Fourier transform (STFT) , the time-frequency resolution couldn't be changed according to the signal characteristics because of the fixed window-length.To solve tbis problem, a window-length-adaptive STFT algorithm based on information entropy is presented.The simulation results indicate that this algorithm can balance the time resalution and frequency resolution based on the signal characteristics effectLvely, and improve accuracy of the instantaneous frequency estimation.%在用短时傅里叶变换(STFT)对全光纤速度干涉仪(AFVISAR)的输出信号进行瞬时频率估计时,由于窗长固定,频谱的时频分辨率不能根据信号特点而改变.为了解决这个问题,提出一种基于最小信息熵准则的自适应窗长STFT算法.仿真结果表明,该算法可以根据信号特征有效地解决时间分辨率和频率分辨率之间的折衷,提高瞬时频率估计的准确性.