Research Papers

Orthogonal Analysis of Multisensor Data Fusion for Improved Quality Control

[+] Author and Article Information
Peng Wang

Department of Mechanical and
Aerospace Engineering,
Case Western Reserve University,
Glennan Building, 10900 Euclid Avenue,
Cleveland, OH 44106-7222
e-mail: Pxw206@case.edu

Zhaoyan Fan

Department of Mechanical,
Industrial and Manufacturing Engineering,
Oregon State University,
Dearborn Hall 217,
Corvallis, OR 97331
e-mail: Zhaoyan.Fan@oregonstate.edu

David O. Kazmer

Fellow ASME
Department of Plastics Engineering,
University of Massachusetts Lowell,
Ball 223, 219 Riverside Street,
Lowell, MA 01854
e-mail: David_Kazmer@uml.edu

Robert X. Gao

Fellow ASME
Department of Mechanical and
Aerospace Engineering,
Case Western Reserve University,
Glennan Building, 10900 Euclid Avenue,
Cleveland, OH 44106-7222
e-mail: Robert.Gao@case.edu

Manuscript received January 31, 2017; final manuscript received May 19, 2017; published online August 24, 2017. Assoc. Editor: Ivan Selesnick.

J. Manuf. Sci. Eng 139(10), 101008 (Aug 24, 2017) (8 pages) Paper No: MANU-17-1065; doi: 10.1115/1.4036907 History: Received January 31, 2017; Revised May 19, 2017

Multisensor data fusion can enable comprehensive representation of manufacturing processes, thereby contributing to improved part quality control. The effectiveness of data fusion depends on the nature of the input data. This paper investigates orthogonality as a measure for the effectiveness of data fusion, with the goal to maximize data correlation with part quality toward manufacturing process control. By decomposing sensor data into a lifted-dimensional space, contribution from each of the sensors for quantifying part quality is revealed by the corresponding projection vector. Performance evaluation using data measured from polymer injection molding confirmed the effectiveness of the developed technique.

Copyright © 2017 by ASME
Your Session has timed out. Please sign back in to continue.



Grahic Jump Location
Fig. 1

Scheme for orthogonal analysis in data fusion

Grahic Jump Location
Fig. 2

Hysteresis loop in pressure–volume–temperature diagram driving changes in molded part dimensions

Grahic Jump Location
Fig. 3

Principal component decomposition in (a) PCA and (b) KPLS

Grahic Jump Location
Fig. 4

Process measurements using commercial sensors (CS) for pressure (P) and temperature (T) and developed multivariate sensor (MVS) for multiphysics measurement within an injection mold cavity

Grahic Jump Location
Fig. 5

Shrinkages predicted using analysis based on specific volume hysteresis of Eqs. (1)(7)

Grahic Jump Location
Fig. 6

Commercial sensors: modeling accuracy and contribution of each sensor

Grahic Jump Location
Fig. 7

Commercial sensors: distributions of projected sensor data with respect to part quality

Grahic Jump Location
Fig. 8

Multivariate sensor: modeling accuracy and contribution from each measured parameter

Grahic Jump Location
Fig. 9

Distribution of projected MVS sensor data points versus part quality




Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In