Predictive-Velocity Modified Particle Swarm Optimization

Velocity update equation of standard particle swarm optimization uses only the size information of previous velocity vector whereas ignoring the corresponding direction information. By adding predictive velocity, a new predictive-velocity modified particle swarm optimization (PVPSO) which using the size and direction information of velocity vectors to estimate and correct particle´s next generation positions is proposed. Theoretical analysis proofs the new algorithm owns an enhanced global search capability. The optimization computing of some examples is made to show that the PVPSO has better global search capacity and rapid convergence speed.