摘要
为抑制单独使用Cohen类进行时-频变换时出现的交叉项,利用经验模态分解法将信号从频域上分离为若干个内禀模态函数之和,将分解后的信号分别进行Cohen类分布的时-频变换,得到信号的时-频分布.对3种不同类型的仿真信号进行计算,并将其时-频分布与直接对信号作Wigner-Ville分布、Cohen类时-频分布(以广义指数核为例)进行比较.结果表明,此方法能够抑制由二次分布所产生的交叉项,得到的结果更加接近理想时-频分布.
To suppress the crossterm interference in the Cohen class quadratic time-frequency distribution,a method based on empirical mode decomposition（EMD） and Cohen class distribution was proposed.In this method,the time-domain signal is first decomposed into a sum of multiple intrinsic mode functions（IMFs） in frequency domain using EMD.Then,the Cohen class distributions of the IMFs are calculated to obtain the sum of all the Cohen class distributions.The time-frequency distributions of three typical simulation signals were calculated by the proposed method,and compared with their Wigner-Ville distribution and Cohen class distribution using generalized exponential kernel.The results show that the proposed method can effectively suppress the crossterms in the quadratic time-frequency distributions,and can produce a more desired time-frequency distribution.
出处
《西南交通大学学报》
EI
CSCD
北大核心
2010年第3期400-404,共5页
Journal of Southwest Jiaotong University
基金
教育部博士点基金新教师资助项目(200806141058)
西南交通大学青年教师科研起步项目资助(2007Q049)
关键词
Cohen类
EMD
交叉项
广义指数核
Cohen class
empirical mode decomposition
crossterms
generalized exponential distribution
