Improved particle swarm optimization algorithm with unidimensional search
Pu Han, Li Meng, Biao Wang, Dongfeng Wang
School of Control and Computer Engineering, North China Electric Power University, Baoding 071003, Hebei, China
In this paper, a strategy of unidimensional search is introduced to particle swarm optimization (PSO). The global exploration capability of PSO is used to identify a promising region in search space. With the region as the starting point, a unidimensional local search is applied to search a more accuracy solution. The local search does not rely on the population information, which makes it can jump out of a local optimum when the population stagnates. With combination of global exploration and local exploitation, the algorithm can discover more favourable search area effectively and obtain a better solution. The improved PSO method is tested on eight benchmark functions. Experimental results show that the method can not only improve the accuracy of solution, but also reduce the influence of initial population distribution upon the algorithm performance. Finally, the influence of parameter variation on algorithm is analysed.