基于蚁群算法的SimuroSot中场队形转换技术  

Ant colony algorithm based on midfield formation conversion system

作　　者：覃杨森[1] 屈四宝[1] 娄云峰[1] 史豪斌[1] 于竹君[1] 

机构地区：[1]西北工业大学计算机学院,西安710072

出　　处：《计算机应用研究》2009年第7期2446-2449,共4页Application Research of Computers

基　　金：国家部委基金资助项目(51315080404);航空科技创新基金资助项目(08E53003);西北工业大学校创新基金资助项目(071069936)

摘　　要：在SimuroSot5v5的中场策略中引入蚁群算法,通过采用信息素分区获取的方式建立起球队中场队形转换机制。该机制以球的位置为驱动信息,根据求取的信息素来确定出适当的中场队形;同时设计任务死锁的自调节模块来提高机器人的自适应能力。仿真实验结果表明,加入以分区方式获取信息素的蚁群算法决策技术,增加了中场队形转换的灵活性,使得球队的整体攻防能力上都得到了相应的改善。In order to improve the flexibility of the game formation of SimuroSotSv5, this article used the way of the information obtained by district-way in the ant colony algorithm, to build a midfield formation of the conversion system. To determine the appropriate midfield formationt, the system used the position of the ball as the drive information, and got the corresponding in- formation elements from the information-distribution chart. At the same time, it improved the ability of self-adaption of each ro- bot, by adding the block of self-regulation module deadlock task. The simulation resuhs show that after added the ant colony al- gorithm decision-making system which gets the information elements by district-way, the football team＇s flexibility of midfield formation is increased, and the team＇s overall defense and offense ability are both be improved accordingly.

关 键 词：机器人足球 队形转换 分区信息素 蚁群算法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]