An Entropy-Based Ant Colony Optimization Algorithm forTraveling Salesman Problems

• Main Authors: Kuo-Sheng Hung, 洪國勝
• Other Authors: Shun-Feng Su
• Format: Others
• Language: en_US
• Published: 2006
• Online Access: http://ndltd.ncl.edu.tw/handle/84525576238430302287

碩士 === 國立臺灣科技大學 === 電機工程系 === 94 === In this thesis, we propose a dynamic updating rule for the heuristic parameters based on entropy to improve the efficiency of ant colony optimization (ACO) in solving the traveling salesman problem (TSP). Our algorithm also proposes to use a lower pheromone trail bound. TSP problems are known as NP-hard problems, which very hard find an optimal solution in a short time. ACO is a new metaheuristic algorithm that has been successfully applied to solve combinatorial optimization problems. ACO algorithm is biologically inspired by one aspect of the behavior of real ants, and it simulates the process of ants searching for food. When ants forage the food, they depend on the amount of pheromone deposited on the traverse path. Although ACO algorithm has very good search capability in optimization problems, it still has some drawbacks such as stagnation behavior, needing longer computing time, and premature convergence. These drawbacks will be more evident when the complexities of the considered problems increase. In our experimental results, the proposed method can avoid stagnation behavior and premature convergence. It can also be found that the proposed dynamic update of the heuristic parameters based on entropy will generate high quality tours and it can guide ants toward the effective solutions space in the initial search stages.