针对正弦调频(SFM)信号Wigner-Ville分布(WVD)存在严重的时频交叉项干扰问题,提出了一种基于稀疏分解的时频分析方法。该方法首先由信号的时频参数构建Gabor原子字典,然后利用匹配追踪(MP)算法实现信号分解,并结合改进遗传算法寻找最佳匹配原子,最后将每次分解得到的Gabor原子通过Wigner-Ville变换叠加得到无交叉项的信号WVD。仿真结果表明,该方法能提高对信号稀疏分解的计算效率,且Gabor原子的选取较为灵活,用少量原子可表示信号WVD。与传统的时频分析方法相比,该方法能有效抑制时频交叉项干扰,且保持高时频分辨率。
For the serious cross- terms interference problem of sinusoidal frequency modulation( SFM) signal Wigner- Ville distribution( WVD), a new time- frequency analysis method based on sparse decomposition is proposed. Firstly, the Gabor atom dictionary is constructed by signal time- frequency parameters, then the signal sparse decomposition is realized by using matching pursuit( MP)algorithm, and combined with the improved genetic algorithm to search the best matching atoms. Finally, the matching Gabor atoms constitutes the signal WVD without cross- terms through the Wigner- Ville transform in turns. Simulation results verify that the proposed method is efficient, it can improve the computational efficiency of signal decomposition, and a few Gabor atoms can represent signal WVD by a flexible selecting way. Compared with traditional time- frequency analysis methods, it can effectively restrain the time- frequency cross- term interference, and maintain a high time- frequency resolution.