Modelling Group-Foraging Behaviour with Particle Swarms

Despite the many features that the behaviour of the standard particle swarm algorithm shares with grouping behaviour in animals (e.g. social attraction and communication between individuals), this biologically inspired technique has been mainly used in classical optimisation problems (i.e. finding the optimal value in a fitness landscape). We present here a novel application for particle swarms: the simulation of group-foraging in animals. Animals looking for food sources are modelled as particles in a swarm moving over an abstract food landscape. The particles are guided to the food by a smell (or aura), which surrounds it and whose intensity is proportional to the amount of food available. The results show that this new extended version of the algorithm produces qualitatively realistic behaviour. For example, the simulation shows the emergence of group-foraging behaviour among particles.

[1]  Jianguo Liu,et al.  Individual-Based Modeling , 2002 .

[2]  Mauro Birattari,et al.  Swarm Intelligence , 2012, Lecture Notes in Computer Science.

[3]  HighWire Press.  Philosophical Transactions of the Royal Society of London , 1781, The London Medical Journal.

[4]  E. P..  Animal Behaviour , 1901, Nature.

[5]  D. Sumpter.  The principles of collective animal behaviour , 2006, Philosophical Transactions of the Royal Society B: Biological Sciences.

[6]  J. Hutchinson.  Animal groups in three dimensions , 1999 .

[7]  Riccardo Poli,et al.  Extending the Particle Swarm Algorithm to Model Animal Foraging Behaviour , 2006, ANTS Workshop.

[8]  R. R. Krausz.  Living in Groups , 2013 .

[9]  Thomas Stützle,et al.  Ant Colony Optimization and Swarm Intelligence , 2008 .

[10]  D. DeAngelis,et al.  Individual-based modeling of ecological and evolutionary processes , 2005 .