Modelling emotion based on fuzzy mathematics and matrix description of finite state machines
-
摘要: 根據大腦的情緒加工環路提出了三個層次的人工情緒框架結構.重點對智能體的底層情緒模型進行了研究,分別采用模糊關系理論和有限狀態機的矩陣形式建立了相應的情緒激活狀態和行為輸出方程.模型考慮了心境和需求對當時情緒的影響,利用矩陣模型可以直接計算出不同情緒狀態下的輸出行為,解決了單純用表的形式記錄"事件—情緒—行為"序列對的存儲空間和查表問題.選擇了嬰兒的情感行為數據來驗證模型的正確性.仿真結果表明:模型在考慮了敏感因子和心境對情緒激活閥值影響的因素后,在機器系統中可以建立有效的情緒與行為輸出模型.Abstract: 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.
-
Key words:
- fuzzy relation /
- finite automaton /
- emotion modelling /
- artificial intelligence
-
點擊查看大圖
計量
- 文章訪問數: 147
- HTML全文瀏覽量: 29
- PDF下載量: 5
- 被引次數: 0
下載:
