文献类型：期刊文献

中文题名：基于非线性函数的非平稳盲源分离步长算法

英文题名：Step-size Algorithm Based on Non-linear Function for Non-stationary Blind Source Separation

作者：李著成[1]

第一作者：李著成

机构：[1]北京联合大学商务学院

第一机构：北京联合大学商务学院

年份：2012

卷号：36

期号：21

起止页码：142-144

中文期刊名：电视技术

外文期刊名：Video Engineering

收录：北大核心:【北大核心2011】；

语种：中文

中文关键词：非平稳盲源分离;步长;收敛速度;稳态误差

外文关键词：non-stationary blind source separation; step-size; convergence speed; steady-state error

摘要：传统自然梯度盲源分离算法采用固定步长,导致收敛速度和稳态误差之间存在矛盾,使算法性能受到制约,特别是对非平稳盲源的分离效果较差。为此,对传统算法进行了改进,提出了一种新的基于非线性函数的步长算法,在分离过程中使步长在每次更新中自动做出合理的调整。通过MATLAB仿真实验证明了新算法的有效性。
The conventional fixed step-size of blind source separation algorithm can＇t be adaptively changed in the process of original signals separation, so both the convergence speed and the steady-state error aren＇t satisfied at the same time, especially for non-stationary signals. To solve the problem, a novel step-size algorithm based on non-linear function is proposed in this paper, and the step-size value can be adaptively changed at each update in the process of separation. The performance is simulated through MATLAB, and the results show that the proposed algorithm has a good performance.