The application of uniform design in parameter establishment of particle swarm optimization

Satisfactory results have been acquired in functional optimization by adopting particle swarm optimization. Its significant feature is simpler expression, less parameters and easier operation. However, selection of key parameters has great influence on algorithm effects. The enactment of the parameters of particle swarm optimization has been determined by experience and experiment. This leads to heavy work and makes the optimal combination of the parameters difficult to obtain. Uniform design theory is applied to the problem of algorithm parameters establishment which can be described as an optimization design problem of multi-factor and multi-level and reduces the work load of experiment greatly of simulation. Simulation results show the robust and effective optimization of the algorithm and indicate the method is feasible and effective.