摘要
针对突现行为产生原因复杂、实际应用广泛的特点,提出一种基于数据挖掘与模糊逻辑的群体突现行为控制方法。首先基于传统Swarm模型,通过遍历8个影响突现行为产生的参数,得到不同参数下的agent的动态数据;同时,利用线性回归和中心聚类算法识别不同参数下所产生的突现行为类型,获得直线型与圆圈型突现行为所对应参数;然后结合C4.5数据挖掘算法,进一步精确参数取值范围;最后,利用模糊逻辑方法实现了多智能体产生圆圈型或直线型突现行为的控制。通过仿真验证了所提出控制方法的有效性。
Aimed at the features that the emergent behavior is widely used and the occurred reasons are complex,a method of group emergent control based on data mining and fuzzy logic is proposed.Firstly, based on the traditional Swarm model,the dynamic data of agents under different parameters are obtained by traversing eight parameters influenced on the occurrence of emergent behavior.Meanwhile,linear regression and central clustering algorithm are used to recognize the type of emergent behavior under different parameters,and obtain the corresponding parameters of linear and circular emergent behavior.Then, the C4.5algorithm is applied to precise parameters'value range.Finally,the control of the occurrence of circular and linear multi-agent emergent behavior is realized by the fuzzy logic method.The effectiveness of the proposed control method is verified by simulation.
作者
朱圣怡
方洋旺
叶圣涛
ZHU Shengyi;FANG Yangwang;YE Shengtao(Aeronautic Engineering College,Air Force Engineering University,Xi'an 710038,China)
出处
《空军工程大学学报（自然科学版）》
CSCD
北大核心
2018年第6期103-108,共6页
Journal of Air Force Engineering University(Natural Science Edition)
基金
2017全国博士后创新人才支持计划(BX201700104)
关键词
突现行为
数据挖掘
模糊逻辑
emergent behavior
data mining
fuzzy logic
