OPTIMIZATION OF SYNCHRONIZATION PARAMETERS FOR HYDROFORMING T-TUBE PROCESS

Abstract

An adaptive, heuristic, nonlinear mathematical model (AHNM) was proposed to optimize the loading path of a hydroforming process as a result of adaptive minimization of the internal pressure and axial load of the process. FEA was used to analyse the process, also this research examined several Machine Learning algorithms such as; Multiple Ridge Regression and Random Forest to learn the relations between the features. The linearity between the features was assumed to create simple AHNM model, where the Multiple Ridge Regression was found to give the highest accuracy. AHNM model was implemented, solved, and optimized using several steps of tee protrusion height. A new Test Rig was developed to experiment the validity of the obtained loading paths for different thicknesses of tube. This research applied the machine learning in this process for the first time, and confirmed that creation of the (AHNM) modelling was successful application.