In the development of smart cities, accurate liquid flow monitoring is essential for the efficient operation of water supply systems. Current flow sensors often face limitations in sensitivity and environmental adaptability, affecting measurement accuracy, and restricting their application in smart city infrastructure. To address these challenges, this study proposes a high-accuracy flow monitoring method. Specifically, by combining the bionic design with advanced signal processing techniques, the sensitivity and anti-interference ability are improved, respectively, to enhance the measurement accuracy. Based on this method, a self-powered flow sensor (SPFS) is developed using noncontact triboelectric nanogenerators (NC-TENGs) as the sensing unit. The SPFS achieves a sensitivity of 2.07 Hz L−1 min−1 and improves the signal-to-noise ratio by more than 13 times over the initial sensing signal. In addition, an intelligent system is developed to accurately measure water resources. The maximum flow rate error rate is less than 0.97% compared to commercial flow sensors. The SPFS demonstrates higher sensitivity and accuracy compared to the existing TENG flow sensors. This study addresses the limitations of existing flow sensors and pioneers a novel solution for enhanced water resource management in smart cities.