CPAG (2024), (стр. 119-126)
АУТОР(И) / AUTHOR(S): Nenad Milosavljevic, Nea Milosavljevic
DOI: 10.46793/CPAG24.119M
САЖЕТАК / ABSTRACT:
Good runnability of paper and board machines is often limited by issues such as web flutter, wrinkles, quality defects, web breaks, and problems with tail threading. When runnability problems are addressed by using unnecessarily high draws, it leads to an increase in web breaks, thereby reducing the machine’s efficiency and productivity. One of the most crucial areas where numerous issues arise is the point where the paper transitions from the top hot cylinder to the bottom cylinder, known as the opening nip position.
To design effective runnability solutions that counteract the negative forces affecting the paper web as it transitions to the drying section, advanced technologies are needed.
Mathematical modeling in the paper industry heavily relies on physical models that include many empirical correlations and assumptions. However, even well-developed mathematical models struggle to accurately describe systems with high degrees of nonlinearity.
This study attempted to simulate the opening nip underpressure using Advanced Neural Network application. The findings indicated a strong alignment between the simulated and experimental data, implying that Advanced Neural Networks could serve as a valuable tool for both designing and diagnosing the performance of existing runnabilty components in paper machines.
КЉУЧНЕ РЕЧИ / KEYWORDS:
Runnablity, neural network, paper machine
ЛИТЕРАТУРА / REFERENCES:
- Levenberg, K., (1944), Quart. Appl. Math. 2, 164-168.
- Marquardt, D., W., (1963), An algorithm for least-squares estimation of nonlinear parameters, J. Soc. Indust. Appl. Math., 11, 431-441.
- Zurada, J., M., (1992), Introduction to Artificial Neural System, West Publishing Company
- Freeman, J., A., Skapura, D., M., (1992), Neural Networks Algorithms, Applications, and Programming Techniques, Addison-Wesley Publishing Company.
- Bulsari, A., B., (1995), Neural Networks for Chemical Engineers, Elsevier Science Publishers, Amsterdam, Netherlands.
- Milosavljevic, N., (2000), New Aspect of Energy Utilization in the Paper Industry, Dr. Thesis, Åbo Akademi University, Åbo, Finland.
- Milosavljevic, N., and Saxen, H., (2001), Hybrid model of cooling tower based on first principle and neural networks, Proceedings of the 5th International Conference on Artificial Neural Networks and Genetic Algorithm (ICANNGA 2001), Praha, Czech Republic.
- Milosavljevic, N., (2009), New Drying and Runnability Technologies for the Increasing Efficiency of Paper and Board Machines, 15th International Symposium in the Field of Pulp, Paper, Packaging and Graphics, June 23rd -26th, 2009, Cigota, Zlatibor, Serbia.
- Leimu, J., (2012), The paper machine cylinder dryer opening nip, Theoretical and experimental investigation, LAP LAMBERT Academic Publishing, Approved by: Turku; Abo Akademi University, Diss., 2008.
- Milosavljevic, N., (2013), Comprehensive approach for the determination of the forces affecting runnability in the single-felted dryer section, 19th INTERNATIONAL SYMPOSIUM IN THE FIELDS OF PULP, PAPER, PACKAGING AND GRAPHICS, June 25 -28th 2013, Cigota, Zlatibor, Serbia, pp. 141-148.
- Milosavljevic, N., (2014), New runnability solution” HiRun P” for the paper and board machines with passive perforated vac rolls concepts in drying section, International Austrian Paper Conference, APV Graz, Austria, June 2014 (another speaker).
- Milosavljevic, N., (2015), The new generation in web stabilization: PrimeRun Evo technology to regulate runnability and save energy, 46°Articelca Congress, The Congress of the Italian Paper Industry, Grand Hotel dei Castelli, Sestri Levante, Genova, Italy, May 28th and 29th, 2015, pp.128-132.