摘要
将拟牛顿方法应用于递归全局最小平均P范数(RTLMP)算法,从而获得了一种适用于α稳定分布噪声的新算法。新算法不仅保持了RTLMP算法的快速收敛性,而且还拥有更加稳定的性能。计算机仿真给出了两种算法的性能比较。
A new algorithm described in this paper can be used in a stable noise environments. This algorithm is obtained by applying the quasi-Newton algorithm to the recursive total least mean P-norm algorithm. It not only keeps fast convergence of the RTLMP algorithm, but also has better stabilization. Performance comparison is indicated by computer simulation.
出处
《制导与引信》
2007年第1期48-52,共5页
Guidance & Fuze
关键词
滤波器
迭代解
递归函数
仿真试验
electric filter
iterative solution
recursive function
simulated test
