摘要
将生物记忆原理与动物行为学原理相结合,利用"联系法"理论,在外部刺激与动物的行为之间建立起映射关系,实现了基于情绪记忆原理的人工鱼行为选择模型。该模型中,人工鱼为出现的每种生物与待选行为分配一个权值;每次行为完成后,人工鱼要对其进行评价,适应生存需求的行为的权值不断增加,不适合的行为的权值将被削弱,直至最终消除。实验表明,该人工鱼模型具有较强的适应性,同时又体现了人工鱼自学习的特性,具有一定的实用价值。
The biological memory principle and animal behaviors theory are combined, the mapping relationship between external stimulation and animal behaviors is established by the relation theory, a behavior selected model of artificial fish based on the emotion memory principle is realized. In the model, an artificial fish assigns a value for every creature and a candidate behavior. After the behavior is finished, the behavior will be evaluated, and the better one’s value is enhanced gradually, the worse one’s value is weakened until being removed. The experiments show that this artificial fish model possesses a better adaptability, implements self-learning character of artificial fish in behavior selected and has better appli-cation value.
出处
《计算机工程与应用》
CSCD
北大核心
2015年第7期45-55,共11页
Computer Engineering and Applications
基金
陕西省重点学科建设专项资金(No.E08001)
陕西省房地产技术经济及管理研究项目(No.E08005)
陕西省科学技术研究发展计划项目(No.2013K11-17)
陕西省教育厅科技计划项目(No.12JK0789)
关键词
人工智能
人工鱼
情绪记忆原理
刺激
artificial intelligence
artificial fish
emotion memory principle
stimulation
