作　　者： ; ; ;

机构地区： 九江学院电子工程学院

出　　处： 《江西师范大学学报（自然科学版）》 2008年第4期432-435,共4页

摘　　要： 为了进一步探索α稳定分布的谱定义,该文提出了广义分数低阶协方差谱的概念,它对α稳定分布随机过程作非线性变换,使变换后的随机过程存在二阶统计量,从而可运用Fourier变换作频域分析.变换函数可以是对数型、Sigmoid型、反正切型等,这些变换函数不依赖于对特征指数α的先验知识的了解或估计,便于工程应用.计算机仿真表明,这些广义分数低阶谱在α稳定分布噪声条件下具有良好韧性,能够对谐波信号频率进行有效识别. Generalized fractional lower-order covariance spectrum is proposed for exploring deeply spectrum definition of stable distribution.After a non-linear transform is used in stable distribution process for getting finite second-order moment,consequently,Fourier transform of transformed stochastic process can be employed to analyze characteristic in frequency domain.Transforming functions are similar to logarithm,sigmoid,arctangent,etc,not depended on a prior knowledge of characteristic exponent,and facilitative for practice application.Simulations show that the defined generalized fractional lower-order covariance spectrum has excellent robust and can identify harmonic frequencies instable distribution process.

关 键 词： 广义分数低阶协方差谱 共变谱 谱

领　　域： [理学] [理学]