Off-Line analysis and prototyping of paper machine drive monitoring system: MODUS-Project Case Study ODT2

Esa Rinta-Runsala

Research output: Book/ReportReport

Abstract

The report concentrates on a specific case study of a single paper machine. Data from the paper machine consists of motor speed and torque measurements taken along the machine as well as supply power and mains voltage measurements. The data is analyzed off-line and some interesting abnormal phenomena in the signals are described. A review of state identification methods is presented, including cluster analysis, decision trees, their fuzzy versions, multi-layer perceptron and radial basis function neural networks. Additionally, self-organizing map (SOM), and adaptive resonance theory (ART) neural networks are presented in more detail and they are prototyped on monitoring of the paper machine. The performance of the prototypes in detecting abnormal phenomena in the data as well as demands and properties of both methods are compared. Finally, a suggestion for implementation is made for using SOM in on-line monitoring. Along the suggestion there are some notions about the limitations of using SOM in monitoring.
Original languageEnglish
Place of PublicationEspoo
PublisherVTT Technical Research Centre of Finland
Number of pages56
Publication statusPublished - 2000
MoE publication typeD4 Published development or research report or study

Publication series

SeriesVTT Information Technology. Research Report
NumberTTE1-2000-27

Keywords

  • SOM
  • self-organizing maps

Fingerprint

Dive into the research topics of 'Off-Line analysis and prototyping of paper machine drive monitoring system: MODUS-Project Case Study ODT2'. Together they form a unique fingerprint.

Cite this