A STUDY ON IMPROVEMENT OF SEARCHING ABILITY OF ANT COLONY OPTIMIZATION
Ant Colony Optimization (ACO) is a new swarm intelligence meta-heuristic which is found to be efficient to solve combinatorial optimization problems. Recently, solution methodsusing ACO for various combinatorial optimization problems, such as the traveling salesman problem, are proposed. The advantage of ACO is to be built a heuristic rule into the algorithms. However, ACO has two weak points that ability of local search is poor and it has no means how to escape from the trap of local minimum solution. Therefore, to improve the search ing ability of ACO, we propose a new ACO with local search and with renewal rule of pheromone values. In this paper, to confirm the searching ability of the improved ACO, Capacitated Vehicle Routing Problem (CVRP) is targeted of the numerical tests. In order to compare the efficiency of the proposed method with ex isting ACO, numerical test has carried out for practical size CVRP. The efficiency of the proposed method has been verified through the numerical test.
ant colony optimization vehicle routing problem combinatorial optimization problem local search
Hidetaka Misawa Shinji Notsu Masakazu Kanezashi
Dept. of Information Processing and Management, Tottori College, Tottori 682- 8555, Japan Facuhy of Engineering, Kinki University, Hiroshima 739 -2116, Japan
国际会议
The Ninth International Conference on Industrial Management(第九届工业管理国际会议 ICIM2008)
日本大阪
英文
783-787
2008-09-16（万方平台首次上网日期，不代表论文的发表时间）