Method for Simultaneously Determining Parameters of at Least One Resin Layer Applied to at Least One Carrier Material

A method for controlling at least one production line, wherein parameters determined in the production line for coating at least one carrier material with at least one resin layer, in particular, at least two, three, or four parameters determined simultaneously, are used for controlling the at least one production line,

• wherein the production line includes:
• at least one Near Infra Red (NIR) multi-meter head for recording at least one NIR spectrum of the resin layer applied to the carrier material in a wavelength range between 500 nm and 2500 nm;
• and
• at least one control system comprising at least one computer aided evaluation unit and a database,
• wherein the at least one NIR multi-meter head is connected to the at least one control system with the at least one computed aided evaluation unit and database for processing and storing the recorded NIR data,
• wherein the evaluation unit is configured to simultaneously determine several desired parameters by an automated comparison of a single NIR spectrum recorded for the resin layer applied to the carrier material with a calibration model created for the respective parameters,
• whereby the database is configured to store parameter data determined in this way in the database,
• where the calibration model is determined using reference samples as follows:
• Recording of at least one NIR spectrum of several reference samples, each with different values of the desired parameters, using at least one NIR multi-measurement head in a wavelength range between 500 nm and 2500 nm;
• Determination of the desired parameters of the measured reference samples using non-spectroscopic methods;
• Assignment of determined parameters to the recorded NIR spectra of the measured reference samples,
• Creation of the calibration model for the relationship between spectral data of the NIR spectra and the corresponding parameter values by means of multivariate data analysis,
• wherein the determined parameters are fed to a “self-learning” AI-based evaluation system (machine learning) for further optimization of the control of the at least one production line and/or for predicting the function of the at least one production line and/or for optimizing the start-up process of the at least one production line.