Êtes-vous un étudiant de l'EPFL à la recherche d'un projet de semestre?
Travaillez avec nous sur des projets en science des données et en visualisation, et déployez votre projet sous forme d'application sur Graph Search.
We present a method to determine the location of an applied pressure on a large area, monolithic silicone based capacitive sensor. In contrast to pressure sensor arrays composed of n x n discrete sensors, we utilize a single sensor body with a single instrumentation interface to detect n pixels. We interrogate the capacitive sensor at different frequencies, thus modulating the effective length of the sensor. These interrogation frequencies are governed by the sensor’s total capacitance, resistance, and desired spatial resolution of the sensor. We developed an analytical model to calculate the frequency response at different length segments of the sensor and used the results to determine the interrogation frequencies for experimental studies. We performed experimental tests on a 1 x n sensor strip and an n x n sensor sheet and showed that we could attain greater than 90% accuracy in predicting the location of the applied pressure using a model generated by a multi-class kernel support vector machine. This approach towards distributed localization of point pressures greatly reduces the hardware complexity in comparison to discrete sensor arrays and increases the physical robustness of the system.
Edoardo Charbon, Francesco Piro, Ashish Sharma
, , ,
Sandro Carrara, Andromachi Tsirou, Amar Kapic