Mathematical Model of Fingertip Skin Under Constant-Current Electrotactile Stimulation

Due to low energy consumption and fast response, electrotactile feedback has shown great potential in human-machine interaction. However, regulating electrotactile perception remains challenging because of the high variability of electrode-skin impedance. Electrode-skin modelling is a common solution, but current researches are still facing many problems such as non-linearity. This paper focuses on voltage response modelling based on data-driven analysis. Two experiments targeting fingertips have been conducted. Significant correlations between pulse amplitude ($PA$), pulse width ($PW$) and peak voltage ($V_{peak}$) ($R^{2}>$0.99) have been found. A mathematical model of fingertip skin is then derived, which enables a precision fitting of the voltage response (RMSE=0.9$\%$). Finally, two calibration methods are proposed for peak voltage prediction. The accuracy (RMSE=2.5$\%$) is also verified under different electrode-skin conditions. The results of this paper are expected to provide novel theoretical support for precise regulation of fingertip electrotactile perception.