PARTICLE SWARM OPTIMIZATION

Particle swarm optimization is a new swarm intelligence optimization technology, which is the simulation of group society. After much research and practice by the PSO, the substance of the PSO can be summed up as “random” and “leadership.” Random is the blood of particle optimization. Particulate optimization is based on the Monte Carlo algorithm of the previous generation, which fully inherits the characteristics of random number resolution. Leadership is the evolution of the particle swarm optimization algorithm. It simulates the phenomenon of mutual cooperation in a group society. Individuals have the ability to learn, and the group has the ability to cooperate with each other. According to the guidance of leader particles, the whole population will converge to the global optimal solution. Therefore, the advantages and disadvantages of PSO are easy to understand. Due to the randomness of PSO, it has advantages in solving nonconvex, discontinuous, high-dimensional, nonlinear, and nondifferentiable optimization problems. Similarly, due to the randomness, its calculation results may be randomness. If the leadership is enhanced, the randomness of the calculation results of the algorithm will be reduced; due to the strong leadership of the PSO, the PSO may easily fall into If the randomness is increased, the algorithm may jump out of the local optimum and find the global optimal.The essence of particle swarm optimization “random” and “leader” have the characteristics of “spear” and “shield,” which need a kind of thought of trade-off to balance.