LAUSR

Sensors based on triboelectric nanogenerators (TENGs) have gained worldwide interest owing to their advantages of low cost and self-powering. However, the detection of most triboelectric vibration sensors (TVS) is restricted to low frequency, whereas high-frequency vibration signals are successfully measured in recent studies; their sensitivity still requires improvement. Hence, a highly sensitive vibration sensor based on TENG (HSVS-TENG) with ultrawide frequency response is presented. This study is the first to introduce a quasi-zero stiffness structure into the TENG to minimize the driving force by optimizing the magnetic induction intensity and the weight of the moving part. The results show that the HSVS-TENG can measure vibrations with frequencies ranging from 2.5 to 4000 Hz, with a sensitivity ranging from 0.32 to 134.9 V g-1 . Moreover, the sensor exhibits a good linear response versus the applied acceleration, and the linearity ranges from 0.08 to 2.81 V g-1 . The self-powered sensor can monitor the running state and fault type of the key components with a recognition accuracy of 98.9% by leveraging machine-learning algorithms. The results reach a new height for the ultrawide frequency response and high sensitivity of the TVS and provide an idea for a follow-up high-resolution TVS.