Exponential Particle Swarm Optimization for Global Optimization

Nature-inspired metaheuristics have been extensively investigated to solve challenging optimization problems. Particle Swarm Optimization (PSO) is one of the most famous nature-inspired algorithms owing its simplicity and ability be used in a wide range applications. This paper presents an extended PSO variant, namely, Exponential (ExPSO). To effectively explore whole search space, proposed algorithm divides swarm population into three equal subpopulations employs new strategy based on exponential function (permitting particles make leaps space) adapted control velocity each particle (to balance exploration exploitation phases). The leaping integrated equation linear decreasing cognitive parameter (including dynamic inertia weight strategy) method. developed allows large jumps at beginning search, then small for further improvements specific regions solution space. Our variant approach, ExPSO, has intensively tested through comparison with eight other well-known heuristic algorithms, over 29 benchmark problems, real engineering Wilcoxon signed-rank test Friedman rank applied analyze performance algorithms. comparisons statistical results show that significantly contributes process proves superiority ExPSO terms convergence accuracy.