自适应人工蜂群算法的盲源分离

Blind source separation based on adaptive artificial bee colony algorithm

盲源分离(BSS)是传感器信号处理领域研究热点,针对传统盲源分离算法大多存在收敛速度慢、分离精度低、适用场合窄的缺点,提出了一种基于自适应人工蜂群算法的盲源分离。利用Givens旋转变换降低计算量,搜索策略引入自适应全局指导项动态调节最优解导向作用,选择策略采用自适应Boltzmann轮盘赌作改进平衡迭代各阶段选择压力集中程度。实验表明:基于自适应人工蜂群算法的盲源分离,能够加快收敛速度并显著提高分离精度至约3个数量级。

Blind source separation（BSS） becomes more attractive targets in sensor signal processing field. A BSS method based on adaptive artificial bee colony algorithm is proposed, aiming at problems of slow convergence, speed low computational precision and narrow application field of existing BSS methods. The algorithm uses Givens rotation to reduce amount of calculation, adaptive global guidance item is introduced in search strategy to dynamically adjust the optimal solution guiding role, and adaptive Bohzmann probability is adopted in selection strategy to adjust selective pressures. Simulation results show that the adaptive algorithm can speed up convergence rate and improve separation precision to about three orders of magnitude.