Nonuniform Temperature Compensation in Ultrasonic Guided Wave Pipeline Health Monitoring Using Local Phase Matching

Abstract

Faced with a complex working environment, pipelines are prone to nonuniform temperature variations, which cause nonuniform phase changes in the guided wave signals during structural health monitoring, thereby increasing the difficulty of monitoring. To address this, a simulation model is established in this paper to analyze the effects of temperature on material parameters and the variation patterns of guided wave signals. A nonuniform temperature compensation method based on local phase matching is proposed. The algorithm first uses cosine similarity to find the locally best-matched signal segments between the monitoring signal and the baseline signal. Then, an indicator is introduced to quantify the differences between these best-matched signal segments, with the maximum difference considered to be the damage index. Three heating experiments on pipelines with nonuniform temperature fields ranging from 24 °C to 80 °C demonstrate that the proposed method can effectively overcome the resulting phase deviations while achieving high detection accuracy and a reduction false positives. Additionally, the method shows high resolution in detecting defects in both temperature-varying and non-temperature-varying regions.