T.B. Chistyakova – Dr.Sc.(Eng.), Head of Department of Computer Design and Control, St. Petersburg State Institute of Technology (Technical University)
M.A. Teterin – Post-graduate Student, Department of Computer Design and Control, St. Petersburg State Institute of Technology (Technical University)
The solution of an actual problem of improvement of quality of products on production of polymeric film materials of carrying out the forecast and operational calculations allowing to investigate most fully cause-and-effect communications of the main technological parameters of production of polymeric films for the purpose of the subsequent issue of the recommendation on production management. The aim of the work is to develop a software package that based on the analysis of large industrial data and based on the calculation of uncontrolled quality indicators of tasks on production indicators allows you to find a set of control actions and factors that ensure the fulfillment of the set values of consumer characteristics of the polymer film and visualize the set of characteristics of industrial production in an ergonomic form for management and production personnel. The core of the software complex is a library of methods of intelligent analysis, including algorithms such as decision trees, neural networks, gradient boosting trees, random forest, logistic regression, algorithm of stochastic embedding neighbors with student distribution (T sne), self-Organizing Kohonen maps. The proposed algorithm for intelligent analysis of big data and quality control of polymer films allows the engineer to issue quality recommendations for production management, to give the trends of the production process to the operator and to issue an order quality card for the management staff. The software package is developed with the use of modern information technologies and is focused on work in various operating systems, including the use of modern mobile devices through the web-interface. The use of the software complex in polymer production can improve the quality of the product resource, significantly reduce the time spent on the analysis of emergency situations in the production, as well as improve the professional level of management and production personnel of plants for the production of polymer films. The software complex has been successfully tested on the data obtained from the calendering lines of plants in Russia and Germany for the production of polymer film materials of Mondi Gronau and Klockner Pentaplast GmbH, and confirmed its performance. The software package can be implemented in foreign and Russian factories, where the quality control of products. The efficiency of the implementation of the software complex is achieved by solving the problem of resource-energy-saving management of polymer films production.
- Kohlert M., König A. Advanced Polymeric Film Production Data Analysis and Process Optimization by Clustering and Classification Methods // Frontiers in Artificial Intelligence and Applications. 2012. 243. 1953−1961.
- Kohlert M., Chistyakova T. Advanced process data analysis and on-line evaluation for computer-aided monitoring in polymer film industry // Journal of Saint Petersburg State Institute of Technology (Technical University). 2015. V. 29. P. 83−88.
- Optical Control Systems GmbH. http://www.ocsgmbh.com.
- Antipin R.V. Avtomatizirovannaya sistema obrabotki informaczii i issledovaniya kachestva polimerny’x plenok. Dis. … kand. texn. nauk. Sankt-Peterburg: SPBGTI. 2008.
- Kohlert M., König A. High Dimensional, Heterogeneous Multi-Sensor Data Analysis Approach for Process Yield Optimization in Polymer Film Industry // Neural Computing and Applications Materials. 2015. 26(3). 581−588.
- Chistyakova T., Teterin M., Razugraev A., Kohlert C. Intellectual analysis system of big industrial data for quality management of polymeric films // Lecture Notes in Computer Science. 2016. T. 9812. P. 234−242.
- Verikas A., Bacauskiene M. Feature selection with neural networks // Pattern Recognition Letters. 2002. 23(11). 1323−1335.
- Seber G., Lee A. Linear regression analysis. John Wiley & Sons. 2003.
- Breiman L. Bagging predictors // Machine Learning. 1996. 24. 123−140.
- Biau G., Devroye L., Lugosi G. Consistency of random forests and other averaging classifiers // Journal of Machine Learning Research. 2008. 9. 2015−2033.
- Breiman L. Random forests. Machine Learning. 2001. 45. 5−32.
- L.J.P. van der Maaten, Hinton G.E. Visualizing High-Dimensional Data Using t SNE // Journal of Machine Learning Research. 2008. 9(Nov). 2579−2605.
- Hinton G.E., Roweis S.T. Stochastic Neighbor Embedding // Advances in Neural Information. Processing Systems. 2002. V. 15. P. 833−840. Cambridge, MA, USA. The MIT Press.
- Chistyakova T.B., Araztaganova A.M., Kolert K. Computer system for thermal shrinkage polymer films obtaining processes control // Vestnik KNITU. 2016. T. 19. № 17. S. 101−105.
- Chistyakova T.B., Polosin A.N., Razy’graev A.S., Kolert K. Programmny’j kompleks dlya upravleniya czvetom tonkix zhestkix polimerny’x materialov // Avtomatizacziya v promy’shlennosti. 2012. № 7. S. 12−18.
- Chistyakova T.B., Kulikov S.I., Polosin A.N., Kolert K. PO dlya upravleniya tolshhinoj kalandrirovanny’x tonkix materialov // Avtomatizacziya v promy’shlennosti. 2007. № 1. S. 8−11.
- Haykin S.S. Neural networks and learning machines. Pearson Education, Upper Saddle River. 2009.
- Segal E., Pe'er D., Regev A., Koller D., Friedman N. Learning module networks // Journal of Machine Learning Research. 2005. V. 6. P. 557−588.
- König A., Gratz A. Advanced Methods for the Analysis of Semiconductor Manufacturing Process Data // Advanced Techniques in Knowledge Discovery and Data Mining. 2005. P. 27−74.