摘要
针对蚁群算法加速收敛和早熟停滞现象的矛盾,根据遗传算法的交叉算子、变异算子和粒子群算法的粒子极值,采用一种优化蚁群算法,以在加速收敛和防止早熟停滞现象之间取得更好的平衡.在利用该算法解决TSP问题中,当前解与个体极值和全局极值分别进行交叉操作,产生的解为新的位置信息.通过对50个城市问题进行实验,结果表明,该方法比一般蚁群算法具有更好的收敛速度和稳定性,适合于求解大规模的问题.
For the conflict of ant colony algorithm for accelerating convergence and premature stagnation phenomenon,according to genetic algorithm crossover operator,mutation operator and particle extreme value of particle swarm algorithm,we studied and improved the ant colony algorithm,to achieve good balance between accelerating convergence and preventing premature stagnation phenomenon.Based on this algorithm,for TSP problem,the current solution is made crossover operation with individual extreme and global extreme respectively.The solution is the new position information produced.The experiment results of 50 urban problems show that it is better than general ant colony algorithm in convergence speed and stability,and suitable for solving large-scale problems.
出处
《微电子学与计算机》
CSCD
北大核心
2011年第8期47-50,共4页
Microelectronics & Computer
基金
天津市科技支撑计划重点项目"基于单目视觉的汽车安全预警技术研究"(10ZCKFGX00300)
关键词
蚁群算法
交叉算子
变异算子
粒子极值
TSP
ant colony algorithm
crossover operator
mutation operator
particle extreme value
TSP
