一种改进的粒子群与人工蜂群融合算法

An improved hybrid algorithm based on particle swarm optimization and artificial bee colony

针对标准的粒子群算法和人工蜂群算法收敛性能差、在复杂优化问题易陷入局部最优的缺点,提出了一种改进的融合算法.改进融合算法拥有双种群并行进化,其中粒子群采用改进的反向学习策略,以增加群体的多样性;蜂群中跟随蜂根据个体停滞次数,自适应地改变进化策略,以平衡全局探索与局部开发能力.同时算法将交替共享两个种群的全局最优位置,通过相互引导使融合算法具有更好的寻优能力.8个经典函数和CEC2013的8个复合函数的实验结果表明,与最新的一些改进粒子群和人工蜂群算法相比,该算法的收敛速度和收敛精度均有较显著的优势.

In order to overcome the shortcomings of standard Particle Swarm Optimization（ PSO） and Artificial Bee Colony Algorithm（ ABC） in complex optimization problems,such as poor convergence performance and easily getting into local minima,an improved hybrid algorithm was introduced. In this algorithm,the population evolves in a double parallel process,in which the particle swarm uses improved opposition-based learning strategy to increase the population diversity,and the employed bees will adaptively change the search strategy according to the number of individual stagnation,balance the global exploration and local development ability. This algorithm will alternate sharing the global optimum of two populations at each iteration; through the guidance of mutual information the hybrid algorithm will get better convergence performance. The experiments are conducted on 8 benchmark functions and 8 composition functions of CEC2013. The result shows that the improved hybrid algorithm performs significantly better than several recently proposed improved algorithm of PSO and ABC in terms of the convergence speed and the solution accuracy.