改进粒子群优化的分段在线盲信号分离算法

发布时间：2018-01-20 07:59

本文关键词： 盲源分离(BSS) 学习速率分阶段学习粒子群优化　出处：《计算机科学与探索》2017年03期 　论文类型：期刊论文

【摘要】：盲源分离(blind source separation,BSS)是指在混合系数未知的情况下,从混合信号中恢复出源信号的过程。在实时盲源分离问题中,学习速率的选择对于算法的性能有着至关重要的作用。为了得到合适的学习速率,提出了如下盲源分离的步长选择算法:通过衡量当前时刻输出信号的依赖程度,将整个信号分离过程分为快速分离和精细分离两个阶段。在快速分离阶段,应用粒子群优化算法确定学习速率,而在精细分离阶段,采用分段函数来确定学习速率。仿真结果证实,新算法比使用固定或其他自适应学习速率的算法有更快的收敛速度和更好的稳态性能。
[Abstract]:Blind source separation (BSS) means that the mixing coefficient is unknown. The process of recovering the source signal from the mixed signal. In the real-time blind source separation problem, the selection of learning rate plays an important role in the performance of the algorithm. A step size selection algorithm for blind source separation is proposed as follows: by measuring the dependence of the output signal at the current time, the whole signal separation process is divided into two stages: fast separation and fine separation. Particle swarm optimization algorithm is used to determine the learning rate, while in the fine separation stage, the learning rate is determined by the piecewise function. The new algorithm has faster convergence rate and better steady-state performance than those with fixed or other adaptive learning rates.
【作者单位】：天津大学电子信息工程学院;天津商业大学信息工程学院;天津大学精密仪器与光电子工程学院;
【基金】：国家自然科学基金No.61401307 国家博士后科学基金No.2014M561184 天津市应用基础与尖端技术研究项目No.15JCYBJC17100~~
【分类号】：TN911.7;TP18
【正文快照】： Grading Learning Based on Improved Particle Swarm Optimization Blind SourceSeparationWANG Zhe1,ZHANG Liyi1,2,CHEN Lei1,2,3+,LI Qiang11.School of Electronic and Information Engineering,Tianjin University,Tianjin 300072,China2.School of Information Enginee

【参考文献】