Juha Kortelainen UPM R&D, Paper and Pulp Finland Avogadro Scale Engineering November 18-19, 2003 The Bartos Theater, MITContentsUPM Key Figures, 2002From the Forest to the CustomerJämsänkoski – Finland, year 2002Jämsänkoski SC PM6Automation Hierarchy, open systemsSlide 8Paper FormationPaper Web Break Camera MonitoringImage analysisOn-line controlTime series data – Multivariate AutoRegressive analysisData ClusteringModelling of paper qualityNeural Networks: Self Organised Maps (T. Kohonen)Clustering of SOM by k-meansSummary for data-amounts / hourJuha KortelainenUPM R&D, Paper and PulpFinlandAvogadro Scale Engineering November 18-19, 2003The Bartos Theater, MITContents●UPM overview●Jämsänkoski Paper Mill●Paper quality and data analysisUPM Key Figures, 2002●One of the world's largest paper producers●Yearly production corresponds to 170,000 km2 area covered by paper! (land area of Massachutes is 20,000 km2)●Mills mainly in Europe, North America and ChinaFrom the Forest to the CustomerJämsänkoski – Finland, year 2002Products: - PM5&6: uncoated magazine 570 000 t/a- PM4: coated magazine 125 000 t/a- PM3: label paper 110 000 t/aFounded: 1888Capacity: 815.000 t/aPersonnel: 940Jämsänkoski SC PM6●325 000 t/a, 39 … 56 g/m², 9.30 m width, 25 m/s speedAutomation Hierarchy, open systemsPaper Formation●micrometer range variations, fibre level−paper surface structure, small defects−optical and printing properties●several meters range, CD and MD profiles−paper web brakes ~ up to 100 km rangePaper Web Break Camera MonitoringImage analysis●Microscopic image analysis for fiber dimensions−fiber length ~2 mm, width ~40 um, cell wall ~ 2 um−automatic fibre analysers with 1,5 um pixel resolution−paper structure with SEM using 0,2 um pixel resolution●Real-time image analysis for web defects and brakes−on-line camera scanner defects down to 0,5 mm size●Real-time microscopic scale?−20 um pixel resolution −10 meter web width−25 m/s speed 12500 images / second with 1 MPix image sizeOn-line control●Distributed Controls−thousands positions●Supervisory Controls: Paper quality data with web scanner−e.g. cross-direction profile control−basis weight−moisture−caliper−colour.:Time series data – Multivariate AutoRegressive analysis●Time dependent cross-correlation disturbance sources●Numerically efficient method needed (FFT)−e.g. 1000 channels, 10 s sample period, 8.6E6 samples/day●Problems:−not efficient enough for long process delays−assumes stationary process state during analysis period−assumes linearity needs data prehandling, about 80 % of manual work!Data Clustering●Automatic clustering often ends up to distinct time periods, which are (more) stationary−product grades, process states●Principal Components, k-means●Neural networks: Self Organised Maps by T.Kohonen−visualization!●Problems:−poor numerical efficiency−does not practically help in data prehandlingModelling of paper quality●Paper strength●Optical properties●PM control variables dominate●some correlation from raw material disturbancesNeural Networks: Self Organised Maps (T. Kohonen)Clustering of SOM by k-meansSummary for data-amounts / hour●DCS data−5 Hz rate, 10,000 channels 2E8 samples / hour−multichannel: vibration, NIR spectra●Paper web scanner−six channels, 1000 Hz 2E7 samples / hour−typically 5 scanners for one production line●Camera systems−many fast speed camera applications in use●off-line image analysis applications real time needs −in future 20 um resolution? 5E13 pixels /
View Full Document