FSS-kernel与FastICA融合的盲源分离算法研究

计算机工程与应用 ›› 2015, Vol. 51 ›› Issue (2): 209-212.

FSS-kernel与FastICA融合的盲源分离算法研究

汪道德1，何鹏举1，龙莉莉2

1.西北工业大学自动化学院，西安 710129
2.湖南科技学院，湖南永州 425000

出版日期:2015-01-15 发布日期:2015-01-12

Novel blind source separation method based on FSS-kernel and FastICA combination

WANG Daode1, HE Pengju1, LONG Lili2

1.School of Automation, Northwestern Polytechnical University, Xi’an 710129, China
2.Hunan University of Science and Engineering, Yongzhou, Hunan 425000, China

Online:2015-01-15 Published:2015-01-12

摘要/Abstract

摘要： FastICA算法有着比传统ICA算法更快、更稳健的收敛速度，但由于其选用的非线性函数不能很好地符合源信号的统计特性，恢复结果并不理想。针对该问题，提出了一种有限支持样本核函数（FSS-kernel）与FastICA融合的盲源分离算法。该方法是通过FSS-kernel算法估计得出源信号概率密度函数，结合FastICA算法，实现混合信号的盲分离。仿真结果表明，该方法能够有效地完成混叠信号的分离，通过与传统ICA算法及FastICA算法比较，证明了该方法具有更高的分离精度和自适应能力。

关键词: 快速独立分量分析（FastICA）算法, 有限支持样本核函数（FSS-kernel）算法, 盲源分离, 算法融合

Abstract: FastICA algorithm has a faster and more robust convergence speed than the traditional ICA algorithm. But its recovery results are not satisfied. Because the specified non-linear function is not well in line with the statistical properties of the source signal. In order to solve the problem, a novel blind source separation method based on FSS-kernel and FastICA combination is proposed in this paper. Probability density function of the source signal is estimated by the FSS-kernel algorithm, and then, to restore the blind separation of mixed signals, FastICA algorithm is used, the negative entropy is the objective function. The simulation results show that the signal aliasing could be separated effectively by this method. It is proved that the method has higher separation accuracy and adaptive capacity, by contrasting with the traditional ICA algorithms and FastICA algorithm.

Key words: Fast Independent Component Analysis（FastICA） algorithm, Finite Support Sample（FSS）-kernel algorithm, blind source separation, algorithm fusion

汪道德1，何鹏举1，龙莉莉2. FSS-kernel与FastICA融合的盲源分离算法研究[J]. 计算机工程与应用, 2015, 51(2): 209-212.

WANG Daode1, HE Pengju1, LONG Lili2. Novel blind source separation method based on FSS-kernel and FastICA combination[J]. Computer Engineering and Applications, 2015, 51(2): 209-212.

[1]	赵青，冶继民，常芳丽. 两正定矩阵联合对角化盲分离算法[J]. 计算机工程与应用, 2019, 55(7): 214-219.
[2]	许德刚，赵萍. 蝙蝠算法研究及应用综述[J]. 计算机工程与应用, 2019, 55(15): 1-12.
[3]	沈成业，张雪英，孙颖，畅江. 小波变换结合盲源分离的EEG情感识别[J]. 计算机工程与应用, 2018, 54(10): 164-168.
[4]	何安玲，何选森. 基于有理非线性函数的Fast-ICA算法[J]. 计算机工程与应用, 2018, 54(1): 251-255.
[5]	陈梦，何选森. 基于八阶收敛牛顿迭代的Fast-ICA改进算法[J]. 计算机工程与应用, 2017, 53(11): 178-181.
[6]	邓大鹏，李骏，丁德强，贺翥祯. 适用于时变系统的在线盲源分离算法研究[J]. 计算机工程与应用, 2016, 52(3): 74-80.
[7]	蔡伟华，何选森. 基于UWT和独立分量分析的含噪盲源分离[J]. 计算机工程与应用, 2016, 52(16): 180-185.
[8]	吴龙华1，朱嘉钢1，陆晓2. 抑制边缘效应的自适应单通道盲源分离[J]. 计算机工程与应用, 2016, 52(14): 130-135.
[9]	蒋照菁，辜方林，张杭. 一种基于NPCA的自适应变步长盲源分离算法[J]. 计算机工程与应用, 2013, 49(8): 206-208.
[10]	陈琛，马庆伦，李灯熬，赵菊敏. NGA实现互信息量最小化的盲源分离[J]. 计算机工程与应用, 2013, 49(4): 121-124.
[11]	董天宝1，程雷2，杨景曙1. 基于时频单源点的欠定混合矩阵估计[J]. 计算机工程与应用, 2013, 49(21): 208-211.
[12]	刘俊良，于凤芹. 基于分离度变步长的自然梯度算法的语音分离[J]. 计算机工程与应用, 2013, 49(21): 216-219.
[13]	徐先峰，刘义艳，段晨东. 基于多步分解算法的解盲源分离新方法[J]. 计算机工程与应用, 2012, 48(30): 162-166.
[14]	郭靖1，2，曾孝平2. 基于Hough变换的时频盲源分离算法[J]. 计算机工程与应用, 2012, 48(21): 26-30.
[15]	董天宝，杨景曙. 一种欠定盲源分离新算法[J]. 计算机工程与应用, 2012, 48(12): 112-115.