Systems Modelling of the Internal Process Variables for Friction Stir Welding Using Genetic Multi-Objective Fuzzy Rule-Based Systems

In order to design and implement a safe and practical Friction Stir Welding (FSW) process, it is crucial to understand the intricate correlations between the controllable process conditions and some internal process variables, such as tool temperature, torque and forces of tool bearing. However, because of the complexity of the FSW process, it is often difficult to derive simple and yet precise enough mathematical models to predict the correlations. In this paper, a systematic data-driven fuzzy modelling approach is developed and employed for the purpose of modelling the internal process properties of FSW, consisting of both the static and dynamic behaviours relating to welding of the AA5083 aluminium alloy. The modelling methodology includes a training data selection mechanism and a hierarchical optimisation structure, which greatly improves the modelling efficiency. The elicited models prove to be accurate and transparent, and they can be used to enhance the welding efficiency and process reliability.