Integrating opposition-based learning into the evolution equation of bare-bones particle swarm optimization

摘要: Bare-bones particle swarm optimization (BPSO) is attractive since it parameter free and easy to implement. However, suffers from premature convergence because of quickly losing diversity, the dimensionality solved problems has great impact on solution accuracy. To overcome these drawbacks, this paper proposes an opposition-based learning (OBL) modified strategy. First, decrease complexity algorithm, OBL not used for population initialization. Second, employed personal best positions (i.e., Pbest) reconstruct Pbest, which helpful enhance speed. Finally, we choose global worst (Gworst) simulates human behavior called rebel item, integrated into evolution equation BPSO help jump out local optima by changing flying direction. The proposed BPSO-OBL, been evaluated a set well-known nonlinear benchmark functions in different dimensional search space, compared with several variants BPSO, PSOs other evolutionary algorithms. Experimental results statistic analysis confirm promising performance BPSO-OBL accuracy speed solving majority functions.