基于信息素改进的蚁群算法  

Ant Colony Algorithm Based on Pheromone Improving

作　　者：孙改平[1,2] 刘春梅[1,2] 

机构地区：[1]北京工业大学,北京100124 [2]华北科技学院,北京东燕郊101601

出　　处：《华北科技学院学报》2010年第1期76-78,共3页Journal of North China Institute of Science and Technology

摘　　要：蚁群算法是一种优秀的启发式算法，具有较强的鲁棒性。针对基本蚁群算法在求解过程中容易出现收敛时间过长以及容易陷入局部最优的不足。本文提出了一种改进的蚁群算法，该算法通过在信息素挥发系数上增加一个收敛函数，加快了收敛速度；通过信息素增量与优秀路径选择相结合，引导算法收敛到最优路径，实验结果表明，改进后的算法在收敛速度和全局寻优能力上有了较大的提高。Ant Colony Algorithm is an excellent heuristic algorithm and has strong robustness. The algorithm easily gets into long convergence time and may be trapped in a loeal optimum. The paper makes some improvement on adding a convergence function to the pheromone volatilization coefficients, improving the convergence rate ; and on guiding the algorithm converges to the optimal path by being combined pheromone increment with the outstanding routing. The experiment results indicate that the improved algo- rithm not only increases the convergence rate and also enhances the ability of searching the global optimal solution.

关 键 词：蚁群算法 信息素挥发系数 信息素增量 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]