1. Particle (agent, individual): each agent in the swarm;
2. Swarm: the population of the agents;
3. Location/Position: agent has n-dimensional coordinates. It represents a best solution for the problem;
4. Generation: each iteration of optimization procedure using the Particle Swarm Optimization to solve problem;
5. Fitness Function: It provides the interface between the optimization problem and the physical problem.
6. Vmax: the maximum velocity value allowed in specified direction.
7. gbest (global best): the position in parameter space of the best fitness value returned to the entire swarm;
8. pbest (particle best): …show more content…
It is capable of evolving towards global optimum solution by its memory mechanism with a random velocity. It has a better performance in searching global optimum solution in a complex search space. Due to its faster convergence rate [20]. Particle Swarm Optimization has a flexible to adjust local and global search abilities. It requires only limited parameters to be adjusted, which makes it more attractive from an implementation viewpoint.
The PSO algorithm has one main operator is the “velocity” equation. It consists of some components in the search space with a velocity (each particle also carries a memory). The velocity provides the search directions for each particle. It can be updated in each iteration of the PSO algorithm. There are three vectors related with each particle in the PSO. They are previous best position current position and velocity.
The PSO algorithm uses a cooperative search approach for optimization. These particles can be interacted with each other. This interaction is achieved by using neighborhoods, wherever a particle can be only interacted with other particles in its neighborhood. The global best (gbest) PSO variants are obtained by the number of neighborhoods to be used [21]. The PSO algorithm performance can be affected by the choice of neighborhood …show more content…
It regulates the maximum step size to the best position of the particle, : is the acceleration constant. It moderates the maximum step size to the global best position in one iteration, are the two random values in the range of [0, 1], w: Inertia weight, : velocity of agent k at iteration I, : current position of agent k at iteration I, : the particle best of agent k and : the global best of the group; The updated velocity equation has three components. They are inertia, cognitive and social component. Inertia component represents as ‘W’. It contains memory of the previous flight direction. According to the velocity equation, the inertia weight ‘w’ is set. (2) where w: tendency of the agent depends on previous velocity for next flight direction, :final weight, : initial weight, :maximum number of iterations and :current number of