Traditional sensors for mechanical equipment monitoring often face challenges such as high cost, difficult installation and removal, and reliance on power sources. Herein, we introduced a polyvinylidene fluoride (PVDF) nanofiber membrane doped with BaTiO3 for constructing a triboelectric acoustic sensor (PB-TAS), innovatively designed for monitoring and recognizing the working conditions of mechanical equipment. This represents the first application of a triboelectric acoustic sensor for mechanical equipment monitoring. The PB-TAS demonstrates an ultra-broad frequency response range of 20–20,000 Hz, effectively covering the characteristic frequency spectrum (within 2000 Hz) during piston pump operation. It also exhibits extraordinarily sensitivity of 4.40 V Pa−1 at 85 dB, and ultra-high frequency resolution down to 0.1 Hz, enabling precise matching of the piston pump's acoustic characteristic frequencies. Integrated with deep learning techniques, the PB-TAS achieves real-time recognition of 15 distinct working conditions of a piston pump, with an impressive accuracy of 95.2 %. The PB-TAS, boasting exceptional sensitivity and frequency resolution, offers a cost-effective, miniaturized, self-powered, non-contact, and highly accurate solution for intelligent monitoring of hydraulic piston pumps.
Read full abstract