Anzeige
Erschienen in:
2022 | OriginalPaper | Buchkapitel
Machine Learning and Real-Time Signal Features Integration for Strength Modelling in Friction Stir Welding Process
verfasst von : B. Das, Jasper Ramon
Erschienen in: Recent Advances in Manufacturing Modelling and Optimization
Verlag: Springer Nature Singapore
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
Abstract
The current research focuses on integrating real-time signal features and artificial intelligent technique to model the tensile strength of the welded joints produced through friction stir welding. Vertical force signals are acquired in real time during the experiments and processed in the time domain to compute the statistical features from the signals. The useful feature domain is generated and integrated with influencing process parameters in the welding process. The data pool created is used for developing the machine learning model to predict joint strength. In the investigation, it is revealed that the integration of signal features in the strength modelling results in better prediction accuracy compared to the result obtained without integration of the signal features in the strength modelling. Strength prediction increases to 99%, with signal features. The proposed method can be effective in an automated platform for effective process monitoring and control in the friction stir welding process.