Recognition of Lateral Driving Scenario for Highway Bridge Monitoring Based on UWFBG

Accurate vehicles lane-level positions of highway bridge are necessary for the construction of the intelligent highway system. This paper proposes a highway bridge vehicle monitoring method based on the ultra-weak fiber Bragg grating (UWFBG) array, which matches the vibration signal distribution to the lateral driving scenario to achieve lane-level monitoring. Firstly, the UWFBG arrays are buried under each lane of the highway bridge to collect vibration signals. Then six common lateral driving scenarios are divided, and the corresponding multivariate time series samples are constructed by the sliding window method. Further, an end-to-end CNN-Bi_LSTM network structure is proposed to extract sensitive features of samples to achieve high-accuracy lateral driving scenario recognition. Experiments show that the proposed method can achieve a recognition accuracy of more than 0.95 in each lateral driving scenario.