Ceramics-Silikáty 34, (4) 343 - 347 (1990) |
|
APPLICATION OF STATISTICAL IDENTIFICATION TO PHYSICAL MODELS OF GLASS FURNACES - PART I
|
Kasa Stanislav 1, Lisý Antonín 1, Vichr Roman 2
|
1 Department of Silicate Technology, Institute of Chemical Technology, Suchbátarova 5,166 28 Prague 6
2 Institute of the Chemistry of Glassy and Ceramic Materials, Czechoslovak Academy of Sciences,
Lipová 5, 120 00 Prague 2
|
The suggested method is based on investigating the dynamic properties of a glass furnace and had the purpose to obtain basic data for adaptive control of furnaces. The procedure simplifies the relationships between the technological quantities and allows random signals to be utilized for changing the input quantities such the electric power input. Determination of the input quantities
and their evaluation is an important part of the statistical identification method. The standard least square method and the extended least square method were used to evaluate the input signals, emploing calculation by the elementary rotation method and the root filter method. The results of statistical identification were verified by means of criteria such as the prediction test, the normal noise test
and the white noise test. |
 PDF (0.5 MB) |
|
