A new method for long-term temperature compensation of structural health monitoring by ultrasonic guided wave

Temperature variations significantly impact ultrasonic guided wave monitoring in non-destructive evaluation and structural health monitoring, particularly in long-distance, irregular waveguides. These variations cause time shifts and amplitude attenuation in guided wave signals, leading to false alarms and missed alarms. To address these challenges, a new temperature compensation algorithm is proposed to improve the stability of ultrasonic guided wave monitoring signals. The algorithm is based on improved optimal baseline selection, the Hilbert transform, and negative exponential amplitude attenuation relationship (IOHN). In the time domain, an incomplete baseline library with a temperature interval of 10 degrees C to 20 degrees C is established, and coarse tuning is carried out through the designed optimal baseline selection strategy to obtain the M = 20-40 most relevant baseline signals; in the Hilbert domain, the average instantaneous parameters of the M signals are designed for fine tuning compensation. Finally, based on the defined compensation index, the robustness and accuracy of the proposed method are verified through high and low temperature experiments, and damage detection experiment, respectively. Experimental results demonstrate the superiority of the IOHN method over traditional approaches. It achieves a maximum residual error reduction of 10 dB and a normalized mean squared error reduction of 0.15 to 0.2 compared to the optimal baseline selection method. In outdoor on-site experiments, it achieves an average recognition accuracy of 99.43 % in detecting simulated damage under a variable temperature environment.