Methods of artificial intelligence for acoustic emission diagnostics of fracture stages (a review)

Based on the analysis of the latest studies, the possibilities of using unsupervised and supervised machine learning algorithms to automate the processing of acoustic emission signals to identify and localize their sources were considered. The accuracy of the results for different approaches was compared, and directions for improvement were described. The importance of further research regarding the adaptation and optimization of the latest techniques for various materials and structures was confirmed.