Multi-Objectivity for Brain-Behavior Evolution of a Physically-Embodied Organism

Jason Teo and Hussein A. Abbass
Artificial Life and Adaptive Robotics (A.L.A.R.) Lab,
School of Computer Science, University of New South Wales,
Australian Defence Force Academy Campus, Canberra, Australia.
{j.teo,h.abbass@adfa.edu.au}

Abstract:

   In this paper, we present a pareto multi-objective approach for evolving the behavior and brain (an artificial neural network (ANN))  of embodied artificial creatures. We will attempt to simultaneously minimize the network size while maximizing horizontal locomotion. A variety of network sizes and behaviors were generated by the pareto approach. The best networks exhibited a higher level of sensory-motor coordination and the creature was able to maintain the walking behavior under different environmental setups.