摘要
自适应陷波器的内部结构使得非二次型误差曲面收敛至局部最优,导致陷波参数停止调整而不能够持续稳定地跟踪信号频率的变化。利用滤波增强信号与原始信号的相关性,设计一个频率跟踪质量评价因子,根据频率跟踪质量可实时监测并自动调整陷波参数。据此原理,提出一种改进的自适应格型陷波频率估计算法。仿真结果与理论分析均表明,该方法在不增加计算量的前提下,较大幅度地提高了算法的收敛速度和持续跟踪精度,证实了该算法优于原有的算法。
An adaptive notch filter＇s inner structure makes a non-quadratic error surface converge to a local minimum, it causes that the parameters of the notch flter stop adjusting so that it cannot track the time-varying frequency of a signal precisely for a long time. Here, the correlation between the enhanced signal and the original signal was used to design a quality evaluation factor, with it the filter＇s parameters could be automatically modified in real-time. According to this idea, a modified frequency estimation method of an adaptive lattice notch filter was proposed. Theoretical analyses and simulation results showed that the convergence of the proposed algorithm and the frequency tracking precision are improved much without additional computations, and the proposed algorithm is better than the existing one.
出处
《振动与冲击》
EI
CSCD
北大核心
2013年第24期28-32,共5页
Journal of Vibration and Shock
基金
国家自然科学基金(61271449)
重庆市自然科学基金重点项目(CSTC
2011BA2015)
关键词
自适应格型陷波器
频率估计
自适应算法
收敛性
adaptive lattice notch filter
frequency estimation
adaptive algorithm
convergence
