摘要
从蚁群觅食行为受到启发,意大利学者M.Dorigo等人提出了一种新型的模拟进化算法———蚁群算法,初步的研究表明该算法具有极强的鲁棒性和发现较好解的能力。该文通过直接模拟真实蚁群的觅食行为,提出了一种真实蚁群模拟算法(RealAntColonySimulatingAlgorithm,RACSA),并通过仿真实验对影响蚁群行为的因素(信息素的重要程度、信息素的蒸发系数、蚂蚁数及信息素留存量)进行了研究,其结论对蚁群算法的理论研究和算法实现具有重要的参考价值。
Inspiration from the foraging behavior of ant colony, a novel simulated evolutionary algorithm called ant colony algorithm (ACA) had been proposed by Italian researcher M. Dorigo el al.. Preliminary study had shown that the ACA algorithm was very robust and had great capabilities in searching better solution. In this paper, a real ant colony simulating algorithm (RACSA) which directly simulates the foraging behavior of ant colony, is proposed. The research on factors such as the importance of pheromone, evaporation coefficient of pheromone, amount of ant and the pheromone quantity deposited by ants is performed by simulated experiments and the results will be guidelines to theoretic research and design of other ant colony algorithms.
出处
《计算机仿真》
CSCD
2004年第8期125-128,共4页
Computer Simulation
关键词
蚁群算法
最短路径问题
组合优化
Ant colony algorithm
Shortest path problem
Combinatorial optimization
