人机交互中的个性化情感模型

Affective model in human-robot interaction

人与机器人的交互过程中,情感因素的引入能够使人机交流更加自然和谐.因此,完整的人工情感模型的建立是首要解决的问题.基于情感能量理论基础,首先,提出了心境自发转移和刺激转移模型.其次,结合情绪自发转移的马尔可夫链模型和刺激转移的HMM模型,将心境和情绪的自发和刺激转移过程统一在一个框架下.最后,将完整的人工情感模型软件化并应用于儿童玩伴机器人上,在接受非结构化环境与用户的信息输入后,个性化的情感软件模块产生输出,实现针对儿童用户的玩伴机器人个性化交互,通过应用验证了该模型的有效性.

Better understanding of affective factors is needed before human-robot interaction （HRI） can be completely natural and harmonious.Establishing a complete HRI affective model is a prerequisite to this understanding.First,based on the theory of affective energy,a mood spontaneous metastasis and stimulus transfer model was proposed.Next,by combining the emotional Markov chain model and the hidden Markov model （HMM）,the spontaneous metastasis and stimulus transfer process of mood and emotion was integrated into a unified framework.Finally,the complete artificial affective models were turned into software modules using Visual C＋＋ and these modules were used to control robotic playmates for children.After receiving non-constructive information from the environment and users,the affective module produced output.In this way the interaction between robotic playmates and children was better managed.