SEMBONE
Parametric Nonlinear Regression Model-based Feedforward control of semiconductor Bonding Equipment
Ever-increasing performance and reliability requirements are being posed on new generations of machines for semiconductor manufacturing, such as wire- and die-bonders, in addition to already advanced functionalities. Advanced motion control is an important enabler of performance and reliability improvements of these machines.
In this project, advanced Multiple-input, Multiple-output (MIMO) feedforward motion control algorithms will be developed for ASMPT semiconductor bonding equipment that compensate for machine behaviors which are usually not captured by the existing feedforwards, such as cable dynamics, frame dynamics, friction, motor driver nonlinearities, etc. These advanced feedforward algorithms could be based on nonlinear regression (NR) models that will serve as augmentations of the existing feedforward control laws. The developed MIMO NR modelling method will be applied in MIMO feedforward motion control of an ASMPT bonding machine, such as wire- or diebonder at the ASMPT Center of Competency in Beuningen.