Modelling and robust controller design for a multi-axis micro-milling machine

Abstract

In the current era of miniaturization, micro manufacturing had became one of the most popular topics. Even tough there are new promising methods such as laser sintering and 3D printing; conventional manufacturing methods continue to hold a unique and irreplaceable position. This thesis aims to design a robust control algorithm for a three axis micro-machining system. In order to synthesize the controller, first the system is modeled. After the modeling, a system identification due to non-linearities is also performed. X, Y and Z axis's identified using prepared Sum of sines identification input. Verification data shows these identified transfer functions represent the physical system well while avoiding over-fit. Using these identified transfer functions, a robust H1 controller is synthesized with designed weighting functions. In simulations, this robust H1 controller showed significantly better performance with or without disturbance. Machining experiments are also done in order to compare the performance of robust controller with the PID controller. According to results of experiments, robust controller showed similar tracking performance with improved surface quality and less oscillations vibrations.