University of Twente Student Theses


Model predictive control of forming processes

Gooijer, B.M. de (2015) Model predictive control of forming processes.

[img] PDF
Abstract:Production standards nowadays are continuously increasing. Therefore there is a call for thorough understanding of production processes and new methods of process control. One of these new methods of process control is by making use of a model based control scheme. The goal of this work is to build a detailed, accurate and fast model of a bending process, which can be used for model-based control. The bending process of interest is the plastic deformation of a 3mm steel flap to an angle of 50°. This flap is part of a demonstrator product especially designed for the MEGaFiT (Manufacturing Error-free Goods at First Time) project in which the University of Twente en Philips cooperate closely. Finite Element (FE) modeling is used to model the bending process. The output of the FE-model is a force curve similar to the force curve as measured in the bending process. Proper Orthogonal Decomposition (POD) is used to reduce the result space from a series of FE analysis. Using a Radial Basis Function (RBF) a meta model is fit through this reduced result space. With an average wall time of 0.03 s, the build PODRBF-model is fast enough to be used in an inverse analysis on which a model based control scheme can be designed. The PODRBF-model is detailed and accurate enough to estimate 3 parameters with 10% accuracy.
Item Type:Essay (Master)
Faculty:ET: Engineering Technology
Subject:52 mechanical engineering
Programme:Mechanical Engineering MSc (60439)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page