Identification and analysis of rutting and fatigue: an optical fibre sensing technology in asphalt pavement

To further enhance the precision of fibre Bragg grating (FBG) sensor data interpretation and intelligently identify rutting and fatigue damage in asphalt pavements, a finite element model of asphalt mixture rutting specimens embedded with high-precision FBG sensors was established first, and a correction formula for the deformation compatibility coefficient of monitoring data was developed. Then, a rutting test and four-point bending experiment were conducted to elucidate the relationship between the sensor-derived monitoring parameters and the pavement disease indicators. Finally, predictive models for the rut depth and fatigue life of asphalt mixtures were developed based on FBG monitoring data, and a methodology for pavement disease identification and analysis was proposed. The results indicate that the deformation compatibility coefficient between the sensor and asphalt mixture was significantly influenced by the number of pavement structural layers and sensor embedding depth but was less affected by the modulus of the asphalt mixture, and the deformation compatibility coefficient can be corrected accordingly. Among the rut depth prediction models, the support vector machine (SVM) model demonstrated superior accuracy and stability compared to multiple regression and BP neural network models, with prediction errors ranging from 2.13% to 2.67%.