Towards a further understanding of the roboticdarwinian PSO

Couceiro, Micael; M. L. Martins, Fernando; Manuel Clemente, Filipe; Rocha, Rui P.; Ferreira, Nuno M. F.

http://hdl.handle.net/10400.26/46928

Use this identifier to reference this record.

Name:	Description:	Size:	Format:
738eb197823864505234eca94271eafa.pdf		250.43 KB	Adobe PDF	Download

Send Feedback

Authors

Couceiro, Micael

M. L. Martins, Fernando

Manuel Clemente, Filipe

Rocha, Rui P.

Ferreira, Nuno M. F.

Abstract(s)

This paper presents a statistical significance analysis of a modified version of the Particle Swarm Optimization (PSO) on groups of simulated robots performing a distributed exploration task, denoted as RDPSO (Robotic DPSO). This work aims to evaluate this novel exploration strategy studying the performance of the algorithm under communication constraints while increasing the population of robots. Experimental results show that there is no linear relationship between the number of robots and the maximum communication range. In general, the decreased performance by the developed algorithm under communication constraints can be overcome by slightly increasing the number of robots as the maximum communication range is decreased.