Experimental PC based TGPID control method for 2D CNC machine

Hendro Nurhadi*, Yeong Shin Tarng

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

3 Citations (Scopus)

Abstract

An important problem in the control of circular motion of CNC machine is to let X and Y axes move simultaneously. This article addresses this problem for the performance of desktop-scale CNC milling machine for reducing roundness error (REB), minimizing position time difference (ΔTt). An approach that can solve those problems will be introduced. Our approach uses a Taguchi-Grey System-Proportional Integral Derivative (TGPID). This method emphasizes an improvement of system performance through this controller's robustness, such as a faster initialization in gaining as appropriate local minima and also high responsive. In this paper, it is aimed to enhance on multi-performance characteristics, namely actual radius (R-act) and position time (Tt). The improvement of roundness error in counter-clockwise (CCW) direction is from 0.151 mm by default, being 0.140 mm by TPID (Taguchi-PID; without grey system), and 0.133 mm by TGPID. The method can reduce the roundness error significantly, also the difference of position time for 100%. This proposed method also offers a simple experimental-based approach. An improvement of its performance indicated that this proposed approach is applied successfully to multi-linear motion performance optimization which is determined by many parameters at multi-quality performances. Performances of the proposed controller scheme, as well as some practical design aspects, are demonstrated by the control of a circular motion of CNC machine.

Original languageEnglish
Pages (from-to)8949-8962
Number of pages14
JournalExpert Systems with Applications
Volume38
Issue number7
DOIs
Publication statusPublished - Jul 2011

Keywords

  • Control
  • Grey
  • Linear motion
  • Optimization
  • PID
  • Taguchi

Fingerprint

Dive into the research topics of 'Experimental PC based TGPID control method for 2D CNC machine'. Together they form a unique fingerprint.

Cite this