Increasing the robustness of distributed genetic algorithms by parallel cooperative-competitive genetic operators
Hernan E. Aguirre, Kiyoshi Tanaka, Shinjiro Oshita · 2001
In this work we study the performance of a distributed GA that incorporates in its core parallel cooperative-competitive genetic operators. A series of controlled experiments are conducted using various large and difficult 0/1 multiple knapsack problems to test the robustness of the distributed GA. Simulation results verify that the proposed distributed GA compared with a canonical distributed GA significantly gains in search speed and convergence reliability with less communication cost for migration. 1