基于细菌群体趋药性的有序盲信号分离算法

摘要

提出了一种基于细菌趋药性的细菌群优化算法.将探测判断策略和优势细菌随机扰动策略引入细菌群体的进化过程,解决了细菌个体在进化后期处于随机摄动状态而难以定位于全局最优位置的问题.通过使用测试函数对算法性能进行测试以及同其他算法的比较表明,所提出的算法在收敛速度和求解质量方面均有很好的性能.进一步将细菌群优化算法应用于解决盲信号分离问题,实现了对源信号的逐一有序盲分离.仿真实验表明,基于细菌群优化算法的盲信号分离算法具有很好的分离效果,可以将源信号按照其规范四阶累积量绝对值的降序进行有效分离.%A bacterial colony optimization algorithm based on bacterial chemotaxis was proposed. Detection determination strategy and random perturbation strategy was led in the process of bacterial colony evolution for solving the difficulty of locating on global optimal position because of the random move state of the bacteria in later period of evolution process. The bacterial colony optimization algorithm proposed has excellent performance in convergence speed and solution quality by comparing with other algorithm using test function. Furthermore, the bacterial colony optimization algorithm proposed was used for solving blind signal separation problem and the sequential blind separation for source signal was realized. Simulation results show that the blind separation algorithm based on bacterial colony optimization algorithm has excellent separation effectiveness. Source signal can be separated efficiently according to the order of their absolute value descending of normalized fourth-order cumulant.