摘要
在有限数据条件下,提出了一种基于数据分段处理和重叠保留的改进循环谱密度估计算法,给出了算法实现的具体步骤,并从改进前后循环谱估计方差、抑制噪声能力和可靠性、分辨率几个方面详细分析了改进算法的性能,推导了相应的表达式。得出了改进算法能够减小循环谱密度估计的方差,在低信噪比时能更有效地抑制噪声对信号循环谱的影响的结论。仿真结果证明了本文所提出算法的有效性。
An innovatory cyclic spectrum estimation algorithm is proposed based on data division and overlap reservation in limited data length.The procedure is given and the performance of the algorithm is analyzed from the point of estimation variance,the ability of restraining noise,resolution and reliability,and the expression is deduced.It is deduced that the innovatory algorithm can reduce the cyclic spectrum estimation variance and the influence of noise on signal’s cyclic spectrum.The simulation results have proved that the validity of the method.
出处
《电路与系统学报》
北大核心
2013年第1期397-402,共6页
Journal of Circuits and Systems
基金
国家自然科学基金(60232010)
国家部委预研资助基金(41001020201)
关键词
有限数据
循环谱
方差
噪声抑制
改进算法
limited data
cyclic spectrum
variance
noise restraining
innovatory algorithm
