A Genetic Algorithm-Based Controller for Decentralized Multi-Agent Robotic Systems

-- In this paper the results of evolution on the task performance of a robot colony are discussed. The cognitive architecture of individual robots of a colony are modified, using genetic algorithms, producing a generation of robots with superior task performance, compared with those of the initial robot population. The effects of mutation probability and fitness scaling parameters on simulated evolution are also studied in this paper.

• Type: Conference
• Year: 1996
• Where: ICEC
• Authors: Arvin Agah, George A. Bekey
• Keywords: ICEC 1996, Robot, Superior Task Performance, Task Performance