Data-based tuning of bumpless feedforward for tracking of multi-phase trajectories with application to wire-bonding machine

Yifan Xu, Silu Chen, John Xu, Chi Zhang, Weizhen Wang, Dunant Halim

Research output: Journal PublicationArticlepeer-review

Abstract

To improve the settling time for high-speed point-to-point motion, a piecewise-model feedforward controller is introduced which utilizes multiple inverse sub-models with bumpless transfer between them. As the transfer function of this bumpless feedforward controller is non-commutative and non-invertible, a set of special perturbation and reference inputs are designed to extract the signals required for computing the gradient of the cost function. In this way, the optimal parameters within different motion phases are found within an integrated process. It is experimentally demonstrated in a wire-bonding machine that the root-mean-square of tracking error is improved by at least 15.3% in settling phase by the proposed data-based tuning method with bumpless feedforward controller, compared with existing tuning methods with uniform feedforward controller.

Original languageEnglish
JournalISA Transactions
DOIs
Publication statusAccepted/In press - 2025

Keywords

  • Bumpless transfer
  • Data-based tuning
  • Feedforward control
  • Gradient
  • Servo control

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Instrumentation
  • Computer Science Applications
  • Electrical and Electronic Engineering
  • Applied Mathematics

Fingerprint

Dive into the research topics of 'Data-based tuning of bumpless feedforward for tracking of multi-phase trajectories with application to wire-bonding machine'. Together they form a unique fingerprint.

Cite this

Xu, Y., Chen, S., Xu, J., Zhang, C., Wang, W., & Halim, D. (Accepted/In press). Data-based tuning of bumpless feedforward for tracking of multi-phase trajectories with application to wire-bonding machine. ISA Transactions. https://doi.org/10.1016/j.isatra.2025.02.028