Convergence Analysis of a Class of Adaptive Ant Colony Algorithm

This paper presents a class of adaptive ant colony optimization algorithm and proves its convergence properties. The global searching and convergence ability are improved by adaptively changing the pheromone trails evaporation factors and decreasing lower pheromone bound. Markov process analysis is used to prove convergence properties of the algorithms. It is shown that its current solutions of the system converge, with probability one, to an optimal solution of the system

[1]  Marco Dorigo,et al.  Ant algorithms and stigmergy , 2000, Future Gener. Comput. Syst..

[2]  G. Di Caro,et al.  Ant colony optimization: a new meta-heuristic , 1999, Proceedings of the 1999 Congress on Evolutionary Computation-CEC99 (Cat. No. 99TH8406).

[3]  Walter J. Gutjahr,et al.  A Graph-based Ant System and its convergence , 2000, Future Gener. Comput. Syst..

[4]  Walter J. Gutjahr,et al.  ACO algorithms with guaranteed convergence to the optimal solution , 2002, Inf. Process. Lett..

[5]  Thomas Stützle,et al.  A short convergence proof for a class of ant colony optimization algorithms , 2002, IEEE Trans. Evol. Comput..

[6]  W. Gutjahr.  A GENERALIZED CONVERGENCE RESULT FOR THE GRAPH-BASED ANT SYSTEM METAHEURISTIC , 2003, Probability in the Engineering and Informational Sciences.

[7]  Marco Dorigo,et al.  Ant Colony Optimization and Stochastic Gradient Descent , 2002, Artificial Life.

[8]  Luca Maria Gambardella,et al.  Ant Algorithms for Discrete Optimization , 1999, Artificial Life.

[9]  Marco Dorigo,et al.  The ant colony optimization meta-heuristic , 1999 .

[10]  Ehl Emile Aarts,et al.  Simulated annealing and Boltzmann machines , 2003 .