Swarm Robot Systems Based on the Evolution of Personality Traits

摘要

Game theory may be very useful in modeling and analyzing swarms of robots. Using game theory in conjunction with traits of personalities, we achieve intelligent swarm robots. Traits of personality are characteristics of each robot that define the robots’ behaviours. The environment is represented as a game and due to the evolution of the traits through a learning process, we show how the robots may react intelligently to changes in the environment. A proof of convergence for the proposed algorithm is offered. The process of selection of traits is discussed and the potential of the modeling is demonstrated in several different simulations.