一种引入自适应动量项的变步长混沌信号盲分离算法

发布时间：2019-02-15 07:43

【摘要】：该文针对混沌信号盲分离问题,提出一种改进盲分离算法。该算法利用信号分离评价指标来构造函数实现步长和动量因子的自适应调整,然后将构造函数代入盲分离算法中并引入自适应动量项。区别于大多数算法不对混合矩阵进行估计的问题,该算法用变步长函数迭代估计出混合矩阵,从而得到全局矩阵和估计评价指标,以此迭代更新步长和动量因子,最终得出分离矩阵。仿真表明,该算法依据估计评价指标构造函数调整步长和动量因子方法是有效的,在平稳和非平稳环境下对混合混沌信号分离时都能达到收敛速度快且稳态误差小的效果;在混入色噪声时,比传统算法抗噪性能好,表明该文算法在混沌信号盲分离处理中有一定应用价值。
[Abstract]:In this paper, an improved blind separation algorithm for chaotic signals is proposed. The algorithm uses the signal separation evaluation index to construct a function to adjust the step size and momentum factor adaptively. Then the constructor is substituted into the blind separation algorithm and the adaptive momentum term is introduced. Different from the problem that most algorithms do not estimate the mixed matrix, the algorithm uses variable step size function to iteratively estimate the mixed matrix, so as to obtain the global matrix and the estimation evaluation index, and then iteratively update the step size and momentum factor. Finally, the separation matrix is obtained. The simulation results show that the algorithm is effective in adjusting step size and momentum factor according to the estimation evaluation index constructor function, and can achieve fast convergence speed and small steady-state error when separating mixed chaotic signals in stationary and non-stationary environment. The anti-noise performance of the proposed algorithm is better than that of the traditional algorithm, which indicates that the proposed algorithm has a certain application value in blind separation of chaotic signals.
【作者单位】：重庆邮电大学信号与信息处理重庆市重点实验室;
【基金】：国家自然科学基金(61671095,61371164,61275099) 信号与信息处理重庆市市级重点实验室建设项目(CSTC2009CA2003) 重庆市教育委员会科研项目(KJ130524,KJ1600427,KJ1600429)~~
【分类号】：TN911.7

【相似文献】