基于一种新的S型函数快速凸组合最小均方算法
发布时间：2018-10-29 12:26
【摘要】：为解决传统凸组合自适应滤波算法在联合参数迭代计算量大、算法收敛速度慢、跟踪性能差等问题,提出了一种基于一种新的S型函数快速凸组合最小均方(SCLMS)算法;该算法用一种新的S型函数,代替Sigmoid函数,在保证和CLMS算法相同稳态误差情况下,避免了指数运算,减少了计算量;同时也提高了收敛速度和信号的跟踪性能。通过独立高斯白噪声作为输入信号算法仿真、相关噪声作为输入信号算法仿真;以及非平稳环境下算法仿真;并对三种仿真结果进行了分析,验证了该算法性能可靠有效。
[Abstract]:In order to solve the problems of the traditional convex combinatorial adaptive filtering algorithm, such as large computation complexity in joint parameter iteration, slow convergence speed and poor tracking performance, a new fast convex combined minimum mean square (SCLMS) algorithm based on S-type function is proposed. In this algorithm, a new S-type function is used to replace the Sigmoid function. Under the same steady state error as the CLMS algorithm, the exponential operation is avoided and the computational complexity is reduced, and the convergence rate and the tracking performance of the signal are also improved. By using independent Gao Si white noise as input signal algorithm simulation, correlation noise as input signal algorithm simulation, and algorithm simulation in non-stationary environment, three simulation results are analyzed, and the performance of the algorithm is verified to be reliable and effective.
【作者单位】： 广西科技大学汽车与交通学院广西汽车零部件与整车技术重点实验室;
【基金】：国家自然科学基金(51665006) 广西高校自然科学基金(2013YB172)资助
【分类号】：TN713
本文编号：2297692
[Abstract]:In order to solve the problems of the traditional convex combinatorial adaptive filtering algorithm, such as large computation complexity in joint parameter iteration, slow convergence speed and poor tracking performance, a new fast convex combined minimum mean square (SCLMS) algorithm based on S-type function is proposed. In this algorithm, a new S-type function is used to replace the Sigmoid function. Under the same steady state error as the CLMS algorithm, the exponential operation is avoided and the computational complexity is reduced, and the convergence rate and the tracking performance of the signal are also improved. By using independent Gao Si white noise as input signal algorithm simulation, correlation noise as input signal algorithm simulation, and algorithm simulation in non-stationary environment, three simulation results are analyzed, and the performance of the algorithm is verified to be reliable and effective.
【作者单位】： 广西科技大学汽车与交通学院广西汽车零部件与整车技术重点实验室;
【基金】：国家自然科学基金(51665006) 广西高校自然科学基金(2013YB172)资助
【分类号】：TN713
【相似文献】
相关期刊论文 前3条
1 谢子殿;董志国;崔师明;;反正切函数联合参数的凸组合最小均方滤波算法[J];黑龙江科技学院学报;2012年06期
2 芮国胜;苗俊;张洋;王林;;基于凸组合的同步长最大均方权值偏差自适应滤波算法[J];通信学报;2012年03期
3 于霞;刘建昌;李鸿儒;;基于箕舌线函数的快速凸组合最小均方算法[J];系统仿真学报;2010年05期
，本文编号：2297692
本文链接：https://www.wllwen.com/kejilunwen/dianzigongchenglunwen/2297692.html
