Particle Swarm Optimization from Theory to Applications

The Basic Of Pso

To understand the PSO, we can imagine a swarm of bees passing through an open field that contains wildflowers. The swarm has a natural desire to locate the highest density position of the flowers in this field. The swarm hasn't any previous knowledge for the field. So, the bees start their search and spread out in random locations. Every bee can remember its locations that have most flowers and this information can be transmitted by communication to the rest of the swarm. With the passage of time, the bees are torn between returning to its previously successful position in finding flowers and heading toward the location (having the most flowers) that reported by the rest of the swarm. The hesitating bee moves in both directions, converting its path to fly somewhere between the two points based on whether social influence dominates its decision or not. Occasionally, a bee may fly over a place in the field that has more flowers than had been discovered by any bee in the swarm. Then, the whole swarm would be drawn as a part in the direction of that location (see Figure 1).

In the figure, dashed lines trace the paths of our imaginary bees, and the solid arrows show their two velocity vector components. The bee No.2 was found the global best position. While, the bee No.1 has found one component of his velocity vector (its personal best location), while the other component is the global best position. The bee No. 3 shows that an example of a particle that have not found a good personal best, it is still drawn to the same position of global best.

Figure 1.

Simulation of PSO by Swarm of Bees Searching for Flowers