Distributed optical fiber sensing technology for wind turbine monitoring

This paper presents a novel method for wind turbine condition monitoring that combines a phase-sensitive optical time-domain reflectometer ($\varphi$-OTDR) with a one-dimensional convolutional neural network (1D-CNN). The $\varphi$-OTDR sensing fiber is laid on the three towers of the wind turbine to collect real-time operating status signals. The acquired signals are processed by the 1D-CNN to perform pattern recognition of the condition signals. Experimental results demonstrate that the recognition accuracy exceeds 95%, proving the effectiveness of our proposed approach.