Multiparameter-based separation method for acoustic emission of in-service prestressed hollow slab tire–road noise signals

Sheng-Li Li,Yi-Qing Zhao,Guang-Ming Wu,Cui-Ping Shi,Nan Jiang,Bin Xu

DOI: https://doi.org/10.1016/j.apacoust.2024.109895

IF: 3.614

2024-02-04

Applied Acoustics

Abstract:The noise caused by tires impact on road surface is inevitable in acoustic emission (AE) monitoring for in-service hollow slab bridges, which affects the damage analysis of the bridge. This gets few attentions due to the lack of such cases of AE monitoring. Since parametric analysis is a commonly used method in AE monitoring, a multiparameter-based method is proposed for separating tire-road AE noise from the mixed AE signal. Firstly, in-site AE acquisition test and bending damage test of hollow slabs were conducted to obtain a mixed signal of tire-road noise and damage. The corresponding AE parameter features were extracted by determining the peak definition time, combined with hierarchical clustering and Laplacian scoring, to select the optimal parameters. The selected parameters were utilized in the k-means clustering procedure to complete the separation, and the final separation effect was verified. Results showed that the mean square frequency, peak frequency, and energy extracted in this study could be considered the optimal parameters for separation of tire–road noise and damage signal. The sequential k-means clustering algorithm could effectively separate the damage signal from the noise signal, which could serve as reference for the study of long-term structural damage monitoring of bridge structures.

acoustics