Machine learning in water machine diagnostics

The thesis focuses on utilizing high-frequency measurement methods for measuring pressure, sound, and acoustic emission, and their application in machine learning for detecting operational states in water machinery. This thesis aims to propose a machine-learning approach capable of evaluating cavitation, flow conditions, and faults based on the collected data. The first part of the thesis discusses the theoretical introduction to selected high-frequency measurement methods and provides a basic introduction to machine learning. The second part describes the experimental setup, including data preprocessing and the implementation of a neural network. In the conclusion, the individual results are evaluated and compared