Fitness feedback based particles swarm optimization

Inertia weight w and acceleration coefficients c are the most effective ways of improving the performance of particle swarm optimization (PSO). A improved PSO was proposed, in which w and c were set to be the function of fitness value and adapted itself in the way of fitness feedback at each iteration. In order to reduce the probability of trapping into a local minimum value, w was recalculated according to the number of iterations, when w equaled to zero during successive M iterations. The proposed adaptive strategy has been implemented and compares with fixed inertia weight PSO (FIWPSO), linearly decreasing inertia weight PSO (LDIWPSO) and nonlinearly decreasing inertia weight PSO (NDIWPSO) employing three global minimum problems. The experimental results establish the supremacy of the proposed variants over the existing ones in terms of convergence speed, robustness and computational precision.