Bridge cable damage identification based on acoustic emission technology: A comprehensive review

Abstract

Bridges have become an essential part of our transport networks. Therefore, the issue of cable damage detection in bridge health monitoring, specifically through acoustic emission (AE) technology, is of paramount importance for the nation and its citizens. This paper extensively investigates and compares various bridge cable damage identification techniques applied to AE. It can be categorized into two major classes: conventional parameter identification and machine learning identification. For the former, the signal’s own feature parameter identification technique and the signal’s mathematical index identification method are reviewed, summarizing their pros and cons. For the latter, both unsupervised and supervised learning are extensively surveyed. This paper introduces insights from AE monitoring in other six research fields, such as rail crack monitoring, and presents a basic process. Finally, the directions and recommendations for future research are proposed, providing useful suggestions for research in bridge cable damage identification and related domains.