传统的信号检测和噪声分离方法只限于在时域或是频域进行，如果信号和背景噪声有很强的时频耦合，就难于在时频域得到好的检测结果。对此提出了分数阶Fourier变换检测算法做出改进，新的算法具有解除时频耦合和抑制噪声的特性，当选择合适的旋转角度使之与处理信号相匹配，就可以在分数阶Fourier域获得较好的检测结果。仿真实验中利用改进的分数阶Fourier变换算法对运动目标进行检测，结果表明新的分数阶Fourier算法的检测性能优于原分数阶Fourier变换。%Traditional signal detection and noise separation methods are restricted in time domain or frequency domain, but the results will not be nice if the signal and noise are coupled intensely. An improved Fractional Fourier transform detection algorithm is presented for this problem. New algorithm can get rid of the coupling and reduce the noise. A preferable detection result will be got in fractional Fourier domain if a suitable rotation angle is adopted to match with the processing signal. Moving target detection is simulated, result shows that the performance of new method detection is better than fractional Fourier transform algorithm.
展开▼
