核递归最小平均P范数算法

发布时间：2018-05-17 05:38

  本文选题：α稳定分布噪声 + 核递归最小平均P范数　； 参考：《信号处理》2017年04期

【摘要】：在强脉冲噪声干扰背景中,核递归最小二乘(Kernel Recursive Least Square,KRLS)算法和核递归最大相关熵(Kernel Recursive Maximum Correntropy,KRMC)算法对非线性信号预测性能严重退化,对此提出一种核递归最小平均P范数(Kernel Recursive Least Mean P-norm,KRLMP)算法。首先运用核方法将输入数据映射到再生核希尔伯特空间(Reproducing Kernnel Hilbert Space,RKHS)。其次基于最小P范数准则和正则化方法,推导得到自适应滤波器的最佳权向量,其降低了非高斯脉冲和样本量少的影响。然后利用矩阵求逆理论,推导得到矩阵的递归公式。最后利用核技巧得到在输入空间高效计算的滤波器输出和算法的迭代公式。α稳定分布噪声背景下Mackey-Glass时间序列预测的仿真结果表明:KRLMP算法与KRLS算法和KRMC算法相比,抗脉冲噪声能力强,鲁棒性好。
[Abstract]:The Kernel Recursive Least Square-KRLS (Kernel Recursive Least Square-KRLS) algorithm and Kernel Recursive Maximum Kernel Recursive Maximum Kernel Recursive Maximum KRMC (Kernel Recursive Maximum Kernel Kernel Recursive Maximum Kernel) algorithm seriously degrade the performance of nonlinear signal prediction in the background of strong impulse noise interference. In this paper, a kernel recursive minimum average P-norm Kernel Recursive Least Mean P-norm-KRLMPalgorithm is proposed. Firstly, the kernel method is used to map the input data to the regenerative kernel Hilbert space. Secondly, based on the minimum P-norm criterion and the regularization method, the optimal weight vector of the adaptive filter is derived, which reduces the influence of the non-Gao Si pulse and the small sample size. Then the recursive formula of matrix is derived by using matrix inverse theory. Finally, the iterative formula of the filter output and algorithm calculated efficiently in the input space is obtained by using the kernel technique. The simulation results of Mackey-Glass time series prediction under the background of 伪 stable distributed noise show that the Mackey-Glass algorithm is compared with the KRLS algorithm and the KRMC algorithm. Strong ability to resist impulse noise and good robustness.
【作者单位】： 杭州电子科技大学通信工程学院;中国电子科技集团第36研究所通信系统信息控制技术国家级重点实验室;
【分类号】：O211.3;TN713
，

本文编号：1900170