A Memetic and Adaptive Continuous Ant Colony Optimization Algorithm

This paper proposes two new variants of the Continuous Ant Colony Optimization algorithm, ACO(R). The first variant, called the Adaptive ACO(R) (AACO(R)), uses the relative diversity of the solutions in the algorithm's archive to adapt its parameters. The second variant, called the memetic AACO(R) (MAACO(R)), uses a local search operator to improve the performance of AACO(R). Both variants were tested on the 22 IEEE CEC 2011 real-world optimization problems and compared with ACO(R) and two state-of-the-art optimization methods. The results demonstrate the merits of the proposed approaches.