Method for process management in paper and cardboard manufacture
AssigneeSiemens Aktiengesellschaft
InventorsHerbert Furumoto, Ulrich Gerdemann, Gerhard Zeiner
Life of this patent
Estimating the term…
Abstract
A method and apparatus for using a neural network to control a paper or paper board production machine. By using spectrum measuring devices, the characteristics of the starting materials for the paper and board production and/or their intermediate or final products are registered and the values fed to a neural network. The network provides statements concerning the paper and board quality, from which signals for the feedback and/or feedforward control of the production process may be derived.
Claims26
A method for the process management of a paper machine for the production of a paper product, the paper machine having a flow box and at least one measuring device for registering characteristic values, and a regulating or controlling device for controlling the paper machine, the method comprising the steps of:
- registering spectral characteristic values at different wavelengths of operating materials of the paper machine using at least one measuring device, the operating materials being at least one of a starting material directly before the flow box of the paper machine, intermediate products of the paper machine or final products of the paper machine;
- evaluating signals from the at least one measuring device with at least one neural network;
- deriving statements about a mechanical quality parameter of the paper product from the evaluated signals;
- and
- deriving control signals from the statements about the mechanical quality parameter of the paper product for feedback control in a stock preparation and/or for feed forward control in the paper machine.
The method according to claim 11, further comprising the step of:
- training the at least one neural network using the quality parameters of the paper or of the board, as measured in a laboratory.
A process for converting a xylem-based material into a finished paper product, comprising the steps of:
- optically scanning the xylem-based material with a spectrometer;
- passing signals representative of the spectrum information obtained by the spectrometer to a neural network;
- analyzing the signals in the neural network to derive information concerning a mechanical quality parameter of the xylem-based material;
- and
- using the derived information to generate signals for controlling the process.
A process for converting a xylem-based material into a paper product, comprising the steps of:
- optically scanning the xylem-based material with a first spectrometer;
- passing optical data obtained by the spectrometer representative of a mechanical quality parameter of the xylem-based material to a neural network;
- analyzing the signals in the neural network to derive information concerning the xylem-based material;
- and
- using the derived information to generate signals for controlling the process.
A method for the process management of a paper machine for the production of a paper product, the paper machine having a flow box, at least one measuring device for registering characteristic values, and a regulating or controlling device for controlling the paper machine, the method comprising the steps of:
- registering spectral characteristic values at different wavelengths of stock suspensions fed to the paper machine using at least one measuring device;
- evaluating signals from the at least one measuring device with at least one neural network;
- deriving statements about the quality of the paper product from the evaluated signals;
- and,
- deriving control signals from the statements about the quality of the paper product for feedback control in a stock preparation and/or for feed forward control in the paper machine.
Dependencies are inferred from the claim text.