Benchmark Tests of Robust Modified Particle Swarm Optimization
Yuanyuan Liu, Wenbo Liu, Ziyang Zhen, Gong Zhang · 2007
The paper presents a new modified approach to improve the global and local exploration capabilities of particle swarm optimization (PSO). The modified PSO is based on the random strategy that random sequences in stead of some difficultly decided parameters are used in the update equation of the particle velocity, in which the inertia weight is replaced by a random sequence and both of two learning rate parameters are replaced by the sum of two different random sequences. Results of comparison with the basic PSO on the examination of some well- known benchmark functions show the perfective and robustness of the improved PSO.