Parameter Adjustment of a Bio-Inspired Coordination Model for Swarm Robotics Using Evolutionary Optimisation.

This work proposed the application of an evolutionary technique to optimise the parameters of a coordination model for swarms of robots. A genetic algorithm with standard characteristics was applied in order to find suitable parameters for the IACA-DI model (Inverted Ant Cellular Automata with Discrete pheromone diffusion and Inertial motion), which, in turn, was proposed in previous works. The IACA-DI is a model to coordinate swarms of robots based on the combination of two bio-inspired techniques: cellular automata and inverted ant system. The main purpose of the model is to carry out surveillance, exploration and foraging tasks. Experiments were performed in different configurations of environments and with different movement strategies to validate this application. Results have shown significant improvements in the model performance compared with previous empirical calibrations, granting a better understanding of the IACA-DI parameters, and allowing significant improvements to be investigated in future works.