Region adaptive scheduling for time-dependent processes with optimal use of machine dynamics

Yanjun Han, Wu Le Zhu, Lei Zhang, Anthony Beaucamp

Research output: Contribution to journalArticlepeer-review


In time-dependent processes, such as bonnet and fluid jet polishing, surface quality and accurate processing critically depend on careful planning of the tool feed. CNC feedrate commands are usually generated from a dwell time map calculated by deconvolution or numerical iteration. These methods are time-consuming, numerically unstable, and fail to consider dynamic stressing of the machine tool. In this research, Gaussian mixture model (GMM) is proposed to model experimental tool influence functions (TIF). This leads to a general analytical convolution model integrating processing depth, volumetric removal rate of TIF, path spacing and feedrate. Based on this model, a novel direct feedrate scheduling method is proposed, which is suitable for any kind of smooth time-dependent processing beam. Optimal feedrate scheduling within dynamic constraints of the machine tool is achieved by establishing acceptable path spacing and feed ranges, whilst dynamic stressing of the machine tool is optimized concurrently through adaptive path spacing. Simulations and experiments demonstrate the enhanced stability and usefulness of the proposed feedrate model in deterministic material removal. It also verifies that path adaptability allows for improved machine tool dynamics, without incurring a process accuracy penalty.

Original languageEnglish
Article number103589
JournalInternational Journal of Machine Tools and Manufacture
Publication statusPublished - 2020 Sept
Externally publishedYes


  • Adaptive path
  • Analytical convolution model
  • Deterministic processing
  • Feedrate scheduling

ASJC Scopus subject areas

  • Mechanical Engineering
  • Industrial and Manufacturing Engineering


Dive into the research topics of 'Region adaptive scheduling for time-dependent processes with optimal use of machine dynamics'. Together they form a unique fingerprint.

Cite this