A novel independent component analysis-based approach for extracting ultrasonic Time-of-Flight in the presence of non-Gaussian co-frequency noise

Ultrasonic non-destructive stress testing plays a pivotal role in the manufacturing of precision equipment. The extraction of the Time-of-Flight (ToF) from an ultrasonic echo signal is critical for accurate stress detection. However, co-frequency noise, which is inevitably present in the echo signal, significantly affects the precise determination of ToF. This study introduces a novel ToF extraction method, referred to as the Co-T algorithm, to mitigate this issue. The proposed method first incorporates a high-frequency signal and a low-frequency signal into the ultrasonic echo signal. Subsequently, the Co-T algorithm is employed to calculate the ToF by evaluating the minima of the auto-correlation results of the signals, which are separated using the Fast Independent Component Analysis (FastICA). Simulation results demonstrate that the proposed algorithm can successfully extract the ToF from signals with a signal-to-noise ratio (SNR) exceeding 0 dB, achieving an identification error of less than 0.04 μs, and demonstrates the robustness against different co-frequency noise types. Furthermore, the proposed method is shown to be superior to existing techniques, including the Hilbert transform, wavelet thresholding algorithms, K-SVD (K-Singular Value Decomposition), and AVMD (Adaptive Variational Mode Decomposition) algorithms. Experimental results validate the efficacy and accuracy of the Co-T algorithm in retrieving ToF from signals in co-frequency noise environments.