Fiber Bragg grating-integrated soil settlement three-dimensional trajectory pipe sensor: dynamic soil subsidence...

Area of Science:

• Geotechnical Engineering
• Sensor Technology
• Machine Learning

Background:

• Traditional soil settlement monitoring methods lack accuracy for dynamic changes, posing risks to engineering projects.
• Accurate, real-time monitoring of soil settlement is crucial for infrastructure safety and stability.

Purpose of the Study:

• To develop an innovative sensor for precise, multi-directional soil settlement strain monitoring.
• To utilize machine learning for predicting loess settlement stages based on sensor data.

Main Methods:

• Development of a fiber Bragg grating-integrated 3D trajectory pipe sensor (SST-3D).
• Experimental validation using air and soil burial tests to identify strain response stages.
• Application of machine learning algorithms (Random Forest) for settlement stage classification and regression.

Main Results:

• FBG wavelength shift demonstrated a linear strain relationship in air tests.
• Soil burial tests revealed four distinct strain response stages during drainage.
• Random Forest achieved 95.65% accuracy in predicting loess collapse stages with a 4.02% regression error.

Conclusions:

• The SST-3D sensor enables precise perception and dynamic monitoring of soil settlement morphology.
• The integrated machine learning approach effectively predicts loess settlement stages, improving safety assessments.