摘要
为在更短的时间内提取舒曼谐振参数,引入对时域采样数据进行直接处理的Prony算法.针对该算法受噪声影响大的缺点,对算法的核心过程———超定方程的求解方法进行了改进,采用Martin等人提出的最小残数法,并与频域处理结果进行对比分析,验证该算法的有效性.
In order to Abstract Schumann Resonance parameters in shorter time,Prony algorithm is adopted and analyzed here to process the sampling data directly without FFT.Due to the fact that the algorithm is noise-sensitive,the minimization of correlated residuals proposed by Martin is adopted for solving the overdetermined equations and verified with simulated signals.It is proved to be feasible for SR signal processing after compared with the results from frequency domain.
出处
《哈尔滨工业大学学报》
EI
CAS
CSCD
北大核心
2012年第7期83-86,91,共5页
Journal of Harbin Institute of Technology
基金
山东省自然科学基金资助项目(ZR2010DM013)
科技部中日合作基金资助项目(2012DFG20510)
关键词
舒曼谐振
PRONY算法
非线性最小平方拟合
谐振频率
最小残数法
schumann resonance
prony algorithm
non-linear least-square fit
resonance frequency
minimization of correlated residuals
