Particle swarm optimiser with hybrid multi-parent crossover and discrete recombination

Dazhi Jiang,Sanyou Zeng,Hui Wang,Zhijian Wu
DOI: https://doi.org/10.1504/IJIIDS.2011.043125
2011-01-01
Abstract:Particle swarm optimiser (PSO) has shown good performance in lots of optimisation problems. However, it easily suffers from premature convergence when solving complex optimisation problems. In order to improve the performance of PSO, this paper presents an enhanced evolutionary algorithm named as PSO with hybrid multi-parent crossover and discrete recombination (PSOHMCDR), which is based on the characteristics of PSO, multi-parent crossover algorithm and differential evolution (DE). Experimental results show that PSOHMCDR outperforms other nine algorithms, including six PSO variants and three typical and effective DE variants.
What problem does this paper attempt to address?