Machine learning coupled RF LC pressure sensing array based on vertical integration structure

RF LC (Radio Frequency Inductance Capacitance) sensors are one of the most promising solutions for flexible, wireless, and passive devices, especially in applications such as electronic skin, human health monitoring, and implantable scenes where complex wiring is not feasible. However, the high-density integration of RF LC sensor in limited area is a significant challenge due to the complex structure and large multi-turn coils in the sensor, which restricts application in implantable and integrated scenarios. In order to enhance the sensitive area and to meet the urgent need for integration in small scale, in this study, we designed a vertically integrated high-sensitivity RF LC pressure sensing array. This array comprises four sensor units, with single-sensor unit area of only 14 mm*14 mm, capable of detecting pressures ranging from 2.5 Pa to 2200kPa, with a maximum sensitivity of 13.875 MHz/kPa. By conveniently changing the turns of the inductance to achieve initial frequency separation in the sensor units. Machine learning methods were introduced to overcome the issues of overlapping resonant peaks and peak disappearance in the S11 spectrum, achieved tactile sensing at nine positions.