Review of Current Trends and Uses of Machine Learning for Discrete Acoustic Emission Interpretation

Abstract

Acoustic Emission (AE) is a well-established and recognised technique for monitoring the degradation of a variety of structures. It is used in a variety of applications, including fatigue monitoring, corrosion monitoring, or detection of pressure leaks. As sensors evolve and databases grow, analysis allows for a better interpretation and understanding of phenomena. Specifically, the usage of Machine Learning (ML) algorithms has proven to be a major tool for interpreting signals. This paper reviews the current usage of ML algorithms used in major Acoustic Emission applications to interpret damage mechanisms, exploring how ML allows the study of more complex phenomena and structures, discussing the conditions, precautions and limitations to its usage as well as future prospects and potentials.