Investigation of Data Augmentation Techniques for Ultrasonic Acoustic Fiber Sensing Signals in Guided Wave-Based Pipeline Damage Detection

This paper examines the efficacy of quasi-distributed acoustic sensors (q-DAS) in identifying damage within pipeline structures, placing a substantial emphasis on generating synthetic q-DAS measurements in active ultrasonic testing setting and bridging the gap between synthetic and real q-DAS measurements. Our research utilizes simulation software to model the ultrasonic guided wave propagation and its interaction with pipeline defects. The pipeline structural health monitoring setup is based on the pulse-echo method utilizing a torsional symmetric mode T(0,1) at 32 kHz, with an aim to identify corrosion and weld irregularities over extensive pipeline lengths. We have prioritized the calibration of simulation models against experimental data, fine-tuning the simulation processes to reflect actual conditions with higher fidelity. The study specifically highlights the simulation's accuracy in capturing the distinct signatures of critical pipeline features and the subsequent detection capabilities within an operational context. By focusing on the experimental validation, we have advanced the understanding and application of structural health monitoring for essential infrastructure, ensuring the simulations' predictive strength aligns closely with real-world sensor data and observed phenomena.