提出了一种对传统粒子群(CPSO)算法进行优化的新型粒子群(GOPSO)算法.与CPSO算法相比,得到可以接受的结果时,GOPSO算法具有更高的收敛速度,需要更少的迭代次数,具有较小的概率陷入局部极值,且易于在实时应用中实现.实验中,已经将GOPSO算法作为高速光纤通信中偏振模色散自适应补偿系统的控制算法.实验数据表明,GOPSO算法的性能大大好于CPSO算法.
A new hybrid particle swarm optimization (PSO) algorithm is presented. The algorithm is named as global best value optimized PSO (GOPS0) algorithm. Comparing with the conventional PSO (CPSO) algorithm, GOPS0 shows faster speed to converge to the global optima, less generation number needed to obtain an acceptable result, lower probability to lock into sub-optima, and easier implementation in the real-world applications. GOPSO has already been used as control algorithm to implement adaptive compensation system for polarization mode dispersion in high-speed optical fiber communication system. GOPSO shows much better performance than that of CPSO.