非线性Volterra系统的总体全解耦自适应滤波

Total Fully Decoupled Adaptive Filter for Nonlinear Volterra System

研究输入、输出观测数据均受噪声干扰时的非线性Volterra系统的全解耦自适应滤波问题.基于总体最小二乘技术和Volterra滤波器的伪线性组合结构,运用约束优化问题的分析方法研究Volterra滤波过程,从而建立了一种总体全解耦自适应滤波算法.并建立了分析该算法收敛性能的参数反馈调整模型,分析表明,该算法可使各阶Volterra核稳定地收敛到真值.仿真实验的结果表明,当输入、输出观测数据均受噪声干扰时,总体全解耦自适应滤波算法的鲁棒抗噪性能和滤波精度均优于全解耦LMS自适应滤波算法.

The fully decoupled adaptive filtering problem of nonlinear Volterra system is investigated with the input and output observation data corrupted by noise. Based on the total least mean square (LMS) technology and the pseudo-linear combination structure of Volterra filter, a total fully decoupled adaptive filtering algorithm is built by using the analysis method of the constrained optimization problem to investigate Volterra filtering process. And the parameter feedback-adjusting model is also built for the convergence analysis of the proposed algorithm. The analysis indicates that the Volterra kernels can evenly converge to the real values by using this algorithm. Simulation results show that the proposed total fully decoupled adaptive filtering algorithm takes on higher robust resistance noise performance and filter precision than the fully decoupled LMS adaptive filtering algorithm, when the input and output observation data are all corrupted by noise.