基于指尖皮肤摩擦的情感规律及模型研究

Emotional intelligence is the key point of bionic robot research. It is significant to realize multi-channel emotional intelligence for medical robots, service robots, disaster relief robots. However, less study focuses on tactile emotion than other senses such as vision and hearing. Tactile perception is mainly evoked by skin friction beside muscle movement. Therefore, this project takes the tactile emotion evoked by skin friction as the research object, and explores the mechanism of the formation and development of the frictional touch emotion. The project will use EEG-fMRI synchrotron imaging technology to build a multi-modal frictional tactile test platform to observe the interaction between the skin and different friction pairs, to discuss the basic characterization of the frictional tactile texture, and to acquire various physiological signals of the human body such as the skin Electricity, heart rate, physiological activity of the brain and other signals. And it aims to reveal the changes of the emotion of the frictional touch, and build the relationship between the characteristics of the object characteristics, skin stress and strain, tactile cognition, and tactile emotion. At last, the emotional model will be established to provide the theoretical basis and technical support for the quantification of tactile emotion. Research achievements have major social value and practical significance in the fields of artificial intelligence, medical and others.

情感智能是仿生机器人研究的热点与难点。多通道的情感智能成为医疗机器人、服务机器人、救灾机器人等亟待解决的问题。而触觉的情感研究明显滞后于视觉、听觉等其他感官。除了肌肉运动触觉，皮肤摩擦触觉是触觉的主要呈现方式。因此，本项目以皮肤摩擦触觉诱发的情感作为研究对象，探讨摩擦触觉情感发生发展形成的机制。项目将利用EEG-fMRI同步联合成像技术，搭建多模态摩擦触觉测试平台，观测皮肤与摩擦副之间的相互作用，讨论摩擦触觉质感的基本表征维度，同时采集人体多种生理信号，如皮肤电、心率、脑生理活动等信号，揭示摩擦触觉的情感变化规律，确立物体特征属性-皮肤应力应变-触觉认知-触觉情感的关系，建立情感模型，为触觉情感的量化提供理论依据和技术支撑。项目研究成果在人工智能、医疗等领域都有重大的社会价值及现实意义。

人机触觉交互是人机社交的重要组成部分，是人机社交安全与舒适的基础。人工智能面临的“情感落地问题”需要首先解决人工智能的触觉交互落地问题。本课题利用多学科交叉的手段，系统地研究了触觉情感环路的各个部分，为触觉的数字化及情感智能提供了理论支持。.设计并制作了触觉刺激呈现装置，以粗糙度为变量设计触觉认知多模态测量实验，通过采集摩擦、振动、脑电及人类真实触感等信号，经过信号处理与分析提取其特征值，并进一步对其进行相关性分析，证明了触觉感知的多模态信息测量方法的可行性，建立了触觉环路上各模态特征之间的关系。搭建了基于EEG-fNIRs同步成像技术的皮肤摩擦触觉多模态试验台，完成了触觉编码及触觉情感的实验研究：第一，采用激光加工了具有不同纹理特征（纹理高度、间距、Hurst指数等）的微观表面作为样本，采集了摩擦、声音、大脑生理活动等完成了的摩擦触觉机理研究实验，探讨了触觉形成机理及编码机制；第二，制作了“自动压力背心”，以模仿人类拥抱，在不同的情绪唤醒程度下，采集了生理体征信号、大脑生理活动，考察了“深接触压力”对情感的舒缓作用。综述了情感模型，建立了基于OCC模型的摩擦触觉情感模型。在摩擦触觉的接触问题上，尝试了基于有限元和光滑粒子动力学耦合算法（SPH-FEM）的手指建模方法，并继续深入开展了基于流固耦合理论的皮肤内触觉感受器的应力应变研究。综述了帕西尼小体结构、功能、力学特性的研究进展，开展了仿生触觉传感器的研究。