Empirical study of the random number parameter setting for particle swarm optimization algorithm

In the particle swarm optimization (PSO) model, the parameters such as the inertia weight have much effect on the performance of PSO and much research has focused on the parameters selection. Because of the programming approach of different high-level programming language and the practice of the programmers, the parameter of the random number is frequently ignored by the researcher. Based on the experiment, the parameter setting of the random number is discussed from the point of view of computer program of PSO. Along with the updating of every velocity vector in one iteration, the random number in the velocity updating equation of PSO model can be set to the same value or the different value for all of the value of every velocity vector, and as a result, different computational results can be evaluated for the same optimization problem. Several standard benchmark functions and the equipment possession quantity optimization model are tested by PSO with the same value and the different value of the random number. The computational results show that the parameter setting of the random number also has important effect on the performance of PSO and PSO with the same value of the random number has much better performance for some unimodal function.