A Centralized & Customized Particle Swarm Optimization and Applications to PID Coefficient Design for the Automatic Voltage Regulator

In this paper, we present a new particle swarm optimization which is inspired by the migration of birds. When we contemplate the migration of the group of birds, there is always a leader bird existed in order to make their fly easier and reduce the whole resistance of air for its own formation of migration. We create a leader particle from choosing the global best position to make all the particles fly closer to the potential best position by moving its own centroid to the present best position that allows all the particles to reshape around it. Another feature of the proposed PSO is that while, by means of the convergence region, the poles of the difference equation of the original PSO system are located in and out of the convergence region randomly, a proposed method has a strict condition that the poles of system always does not lie outside and tend to be in the convergence region. To satisfy the condition, the inertia is a function of the cognitive and social coefficients. With the proposed PSO algorithm we design PID coefficients for the automatic voltage regulators. We show that the proposed algorithm has a far better performance than the original PSO for the PID design of automatic voltage regulators.