摘要
针对粒子群优化(particle swarm optimization,PSO)算法收敛速度慢、易陷入局部最优等缺点,提出了一种基于改进PSO算法优化的独立分量分析(independent component analysis,ICA)算法,用于水下语音和噪声混叠信号的盲分离.采用规范四阶累积量的绝对值作为ICA中的目标函数,通过改变惯性因子ω和压缩因子k来增强粒子的自适应寻优能力.对比实验结果表明,该算法在收敛速度、算法稳定性和分离效果方面的性能更优.
A new independent component analysis （ICA） algorithm optimized from the improved particle swarm optimization （PSO） is proposed to overcome the drawbacks of the slow convergence speed and the aptness into local minimum of the PSO algorithm. The proposed method is aimed at extracting the target speech signal in the under-water noisy environment. It uses the absolute value of normalized fourth-order cumulant as an objective function. By changing the inertia factor ω and constriction factor k, particles have more adaptive ability to find out the optimal particle quickly. Comparing with the classical PSO algorithm, the proposed improved method performs faster convergence speed, better algorithm stability and superior separation effect.
作者
王光艳
耿艳香
陈雷
WANG Guang-yan;GENG Yan-xiang;CHEN Lei(College of Information Engineering,Tianjin University of Commerce,Tianjin 300133,China)
出处
《应用科学学报》
CAS
CSCD
北大核心
2018年第4期589-600,共12页
Journal of Applied Sciences
基金
国家自然科学基金青年科学基金(No.6140010551)
天津市应用基础与前沿技术研究计划重点项目基金(No.14JCZDJC32600)
天津市大学生创新创业训练计划项目基金(No.201710069052)资助
关键词
水下语音通信
盲语音分离
独立分量分析
四阶累积量
粒子群优化
under-water speech conmmnication
blind speech separation
independent component analysis （ICA）
fourth-order cunmlant
particle swarm optimization （PSO）
