Crack-Sensitive Multipoint FBG Tube Sensor Subsurface Deformation Monitoring Tube for Real-Time Measurement of Soil Layer Axial Deformation

To address the limitations of conventional methods in terms of accuracy and applicability for deep soil displacement monitoring, this study proposes a crack-sensitive multi-point fiber Bragg grating (FBG) tube sensor. Applying external loads of different thicknesses on the sensing tube to simulate the displacement of underground soil, this sensor can achieve real-time monitoring of the axial deformation and displacement evolution within the soil layer. The results indicate that this sensor exhibites stable strain transfer characteristics under different overburden conditions, following a pattern in which strain (displacement) gradually decreased from the pulling end toward the fixed end. Moreover, the sensitivity of the cracked tube is significantly higher than that of the intact tube, highlighting the advantage of the crack-sensitive design. In addition, to improve the accuracy of strain–displacement data interpretation, an XGBoost-based modeling approach is introduced to predict and validate experimental results, achieving an average prediction error of 0.229 mm with a regression prediction success rate of 99.79%. This performance corresponds to an expanded displacement measurement uncertainty of approximately 1.0 mm at an approximate 95% confidence level, demonstrating the reliability of the proposed sensing and data-driven approach. Further analysis based on the confusion matrix confirmed that the method achieved highly reliable classification of displacement responses under different overburden thicknesses, with almost no misclassification. In summary, the proposed crack-sensitive multi-point FBG tube sensor exhibites the significant advantages in displacement sensitivity, monitoring stability, and predictive accuracy, which shows strong potential for application in deep displacement monitoring of slopes and roadway infrastructures, offering a promising pathway for soil stability assessment and disaster early warning.