The Influence of Topologies on the Dynamic Vector Evaluated Particle Swarm Optimisation Algorithm

Most real world problems have more than one objective, with at least two objectives in conflict with one another and at least one objective that is dynamic in nature. The dynamic vector evaluated particle swarm optimisation (DVEPSO) algorithm is a co-operative algorithm, where each sub-swarm solves only one objective function and therefore, each sub-swarm optimises only a sub-set of decision variables. Knowledge is shared amongst the sub-swarms when the particles' velocity is updated, by using the position of the global guide of the sub-swarm or of another sub-swarm. Each sub-swarm's entities are connected to one another according to a specific topology that determines the communication of particles with one another. This paper investigates the effect of using the star or Von Neumann topology for DVEPSO's sub-swarm's particles. The results indicate that the star topology performed the best with regards to accuracy and the Von Neumann topology performed the best with regards to stability. In addition, the Von Neumann topology performed the best on benchmarks with a non-linear Pareto-optimal set (POS) and in very fast changing environments.