Abstract
Double-row tapered roller bearings (DTRBs) are widely used in wind turbines because of their high load-bearing capacity and durability. However, wind turbines typically operate in harsh environments, subjecting bearings to complex working conditions, which significantly increases the difficulty of operational status monitoring. Traditional monitoring methods rely on external power sources and complex sensor networks, which make them susceptible to environmental interference, and complicated to maintain. This paper presents an innovative, integrated symmetrical single-electrode triboelectric double-row tapered roller bearing (SST-DTRB) by incorporating a triboelectric nanogenerator (TENG) with DTRB. This scheme converts the frictional energy generated during bearing operation into electrical output, producing signals that enable simultaneous sensing of both ends of DTRB. Experimental results demonstrate that this monitoring scheme exhibits high sensitivity, stability, and reliability, with excellent robustness in material selection and design gap, and is capable of long-term operation without external power sources. The effectiveness and self-sensing capability of SST-DTRB under variable speeds are validated using a wind turbine test bench. High-accuracy bearing fault diagnosis under multiple conditions is achieved based on time-frequency transformation and deep residual neural networks. The proposed SST-DTRB provides in situ self-powered sensing capability for wind turbines and offers new insights in the development of intelligent sensing systems.