蚁群智能体记忆模型研究-计算机工程与应用.PDF

46 2010 ，46 （27 ） Computer Engineering and Applications 计算机工程与应用 蚁群智能体记忆模型研究 黄光球，邢玉飞，赵 煜 HUANG Guang-qiu ，XING Yu-fei ，ZHAO Yu 西安建筑科技大学 管理学院，西安 710055 School of Management ，Xi ’an University of Architecture and Technology ，Xi ’an 710055 ，China E-mail ：huangnan93@ HUANG Guang-qiu ，XING Yu-fei ，ZHAO Yu.Research on memory model of ant colony agent.Computer Engineering and Applications ，2010 ，46 （27 ）：46-51. Abstract ：To introduce classical ant colony algorithm into memory mechanisms ，regards ant as the agent with memory ，and by analyzing the principles of storage ，update and forget ，a model of ant colony agent based on biological memory principles is es- tablished.In the model ，ant agents can constantly search and update its own and ant colony memory banks and adjust current optimal path immediately by virtue of memory.The combination of memory principles and classical ant colony algorithm have taken the place of latter ’s optimal pattern of iterative repetition ，which can realize path optima much better ，shorten the search- ing time and make the algorithm more efficient.The application shows that the model can track the latest development in real time ，adjust and judge the current optimal path immediately ，thus guide it and other ant agents to find the optimum path. Key words ：memory principle ；ant colony algorithm ；ant colony agent ；memory bank ；path optima 摘 要：将记忆机制引入传统蚁群算法，把蚂蚁看作具有记忆的智能体，通过对记忆的存储、更新及遗忘原理进行分析，建立一种 基于生物记忆原理的蚁群智能体记忆模型。在模型中，蚂蚁智能体能够凭借记忆对自身记忆库和蚁群记忆库进行不断搜索和更 新，及时调整当前最优路径。记忆原理与传统蚁群算法的结合替代了后者多次反复迭代的寻优模式，能更好地实现路径选优、缩 短搜索时间、提高算法执行效率。实践检验发现，该模型能实时追踪蚁群智能体的最新动态，对当前最优路径做出及时调整和判 断，从而引导自身及其他蚂蚁智能体准确、高效地寻找到最优路径。 关键词：记忆原理；蚁群算法；蚁群智能体；记忆库；路径选优 DOI ：10.3778/j.issn. 1002-8331.2010.27.012 文章编号：1002-8331（2010 ）27-0046-06 文献标识码：Ａ 中图分类号：TP 18 [4] 蚁群算法（Ant Colony Algorithm