模糊数学和有限状态机矩阵形式描述的人工情绪模型  

Modelling emotion based on fuzzy mathematics and matrix description of finite state machines

作　　者：史雪飞[1] 王志良[1] 张琼[1] 

机构地区：[1]北京科技大学信息工程学院,北京100083

出　　处：《北京科技大学学报》2010年第9期1238-1242,共5页Journal of University of Science and Technology Beijing

基　　金：国家高技术研究发展计划资助项目(2007AA04Z218);北京市自然科学基金重点项目(KZ200810028016)

摘　　要：根据大脑的情绪加工环路提出了三个层次的人工情绪框架结构.重点对智能体的底层情绪模型进行了研究,分别采用模糊关系理论和有限状态机的矩阵形式建立了相应的情绪激活状态和行为输出方程.模型考虑了心境和需求对当时情绪的影响,利用矩阵模型可以直接计算出不同情绪状态下的输出行为,解决了单纯用表的形式记录"事件—情绪—行为"序列对的存储空间和查表问题.选择了婴儿的情感行为数据来验证模型的正确性.仿真结果表明:模型在考虑了敏感因子和心境对情绪激活阀值影响的因素后,在机器系统中可以建立有效的情绪与行为输出模型.A three-layer emotion model structure based on brain science was proposed firstly.Then a bottom-level emotion model of the intelligent system was presented in detail.It was constructed by using fuzzy mathematics and matrix description of finite state machines.A fuzzy relation between stimulus and emotion was produced to determine the active emotion in consideration of mood and desire at that time.The output behavior was calculated by a matrix style of finite state machines while the active emotion was given.The model constructed in this way could reduce memory spaces used only for storing the corresponding relations among stimulus,emotion and behavior.The simulation result and conclusion are presented in the end.

关 键 词：模糊关系 有限自动机 情绪模型 人工智能 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]