PS2O: A multi-swarm optimizer for discrete optimization
Hanning Chen, Yunlong Zhu, Kunyuan Hu, Tao Ku · 2008
In this paper, we implement an entire social system which consists of both heterogeneous cooperation and homogeneous cooperation aspects to formulate our simulation models of coevolution. We introduced a number of N species each possesses a number of M individuals into this coevolution model to represents the “biological community”. Each individual of the community evolves based on the knowledge integration of itself, its species member and its symbiotic partners from other species. Since the community is made up of a swarm of agents who are species while each species is made up of a swarm of species members (individuals), our swarms within swarm model is instantiated as a hierarchical coevolutionary optimization algorithm, namely Particle Swarms Swarm Optimizer (PS2O). The PS2O algorithm is evaluated on four discrete optimization problems for compared with the canonical discrete PSO algorithm. The comparisons show that on average, PS2O outperforms the PSO in terms of accuracy and convergence speed on all benchmark functions.