电触觉的皮肤神经机理仿真及实验验证

触觉学已经在很多领域得到应用，如遥控机器人、手术机器人、假肢、娱乐交互界面以及虚拟现实技术等，以此来增强人对机器或者虚拟物体的操作性。电触觉是现在主要的触觉再现措施，即通过改变恒流/恒压电脉冲的频率、脉宽、强度以及脉冲方向等因素,让人产生不同的触觉感。首先通过人体皮肤神经的建模推导出激励函数，并进行仿真研究，通过改变施加在电刺激阵列的电脉冲方向和强度，设计能分别刺激皮下3种刺激感受器（Meissner触觉小体、Merkel触盘、Pancinian环层小体）的实验范式，同时进行了心理物理学实验。10位受试者参加了实验，对食指进行电触觉刺激。正脉冲时采用不同频率（10、30、70、90 Hz）的电脉冲进行刺激，让受试者产生不同级别的振动感；负脉冲时采用不同脉宽（150、200、250、300 μs）的电脉冲进行刺激，让受试者产生不同级别的压力感。受试者对振动和压力的感觉强度进行主观判断。统计结果表明,该模型下的实验模式能够使触觉强度分级平均识别率高达80%以上，同时可通过对刺激电极分布以及电流大小的分析，找到最佳的刺激模式，保证最佳的触觉再现。

Stimulation and Experimental Verification on Cutaneous NervousMechanism of Electrotactile Feeling

Nowadays, haptics has been widely used in many fields such as telecontrol robots, surgical robots, prosthesis, entertainment interface and virtual reality, which has improved the operability of actual or virtual objects for human. Electrotactile stimulation is a main technique to reproduce tactile sensation by changing the frequency, duration, amplitude, direction and distribution of constant current electrical pulses. In this article, a model of cutaneous nerves under electrical stimulation was developed to serve for simulation study, and stimulation function was deduced based on the model. With the simulation, optimal paradigms for stimulating three mechanoreceptors (Meissner’s corpuscle, Merkel’s disc and Pancinian corpuscle) were separately found in a trial and error way, i.e. by changing the direction and amplitude of electrical pulses in arrays through simulation study. Furthermore, psychophysical experiments were carried out to make 10 experimenters sense two kinds of feelings (vibration and pressure). The targeting area for electrotactile stimulation was on the index fingers. For vibration feeling, positive electrical pulses of different frequencies（10，30，70，90 Hz）were applied; For pressure feeling, negative electrical pulses of different widths（150、200、250、300 μs）were applied. The subjects gave their subjective judgment on the grade of the feeling generated by electrical stimulation.Statistical results showed that the averaged recognition rate of two sensations under the help of the model  reached higher than 80%. The model is helpful in finding the best stimulation pattern and assuring the realization of optimal reproduction of tactile sensation.