Estimation of Extrusion Process Parameters in Tire Manufacturing Industry using Random Forest Classifier

Abstract

The extrusion process is a very complex process due to the number of process parameters involved. Throughout the workflow process, the process parameters are determined by trial-and-error method according to the recipe of materials. This technique causes loss of production and time as well as energy consumption. In extrusion, temperature and speed parameters are very important to obtain a homogeneous raw material product input and high-quality extruded products. It is necessary to monitor the temperature changes and process speed control during the flow of the molten raw material between the barrels of the extruder machine, which is the extrusion equipment. By monitoring the extruder in real time, estimating the extrusion process parameters according to the amount of product to be produced will make the extrusion process operations more efficient. In this study, a classification algorithm to process these parameters is developed in the “Pycharm” environment and the model is trained with the supervised learning method using the image processing algorithm outputs. The model is able to estimate the extruder 'speed and temperature parameters' and the 'ready to run' decision of the machine with 93% success for different production quantities entered by the operator.

Keywords

• Extrusion process
• tire manufacturing
• decision support estimation algorithm
• surface inspection
• random forest algorithm

References

1. [1] V. García, J.S. Sánchez, L.A. Rodríguez-Picón, L.C. Méndez-González, H.D.J. Ochoa-Domínguez. "Using regression models for predicting the product quality in a tubing extrusion process." Journal of Intelligent Manufacturing, vol. 30, 6, 2019, pp. 2535-2544.
2. [2] C.Y. Wu, Y.C. Hsu. "Optimal shape design of an extrusion die using polynomial networks and genetic algorithms." The International Journal of Advanced Manufacturing Technology, vol. 19, 2, 2002, pp. 79-87.
3. [3] S.A. Oke, A.O. Johnson, O.E. Charles-Owaba, F.A. Oyawale, I.O. Popoola. "A neuro-fuzzy linguistic approach in optimizing the flow rate of a plastic extruder process." International Journal of Science & Technology, vol. 1, 2, 2006, pp. 115-123.
4. [4] R.S. Sharma, V. Upadhyay, K.H. Raj. "Neuro-fuzzy modeling of hot extrusion process." Indian Journal of Engineering & Materials Sciences, vol. 16, 2009, pp. 86-92.
5. [5] S.H. Hsiang, Y.W. Lin, J.W. Lai. "Application of fuzzy-based Taguchi method to the optimization of extrusion of magnesium alloy bicycle carriers." Journal of Intelligent Manufacturing, vol. 23, 3, 2012, pp. 629-638.
6. [6] A. Chondronasios, I. Popov, I. Jordanov. "Feature selection for surface defect classification of extruded aluminum profiles." The International Journal of Advanced Manufacturing Technology, vol. 83, 1, 2016, pp. 33-41.
7. [7] S. Ravi, P.A. Balakrishnan. "Temperature response control of plastic extrusion plant using MATLAB/Simulink." International J. of Recent Trends in Engineering and Technology, vol. 3, 4, 2010, pp. 135-140.
8. [8] URL: https://www.substech.com/dokuwiki/doku.php?id=extrusion_of polymers (Access: Jan 9, 2023).

9. [9] Y. Mishina, R. Murata, Y. Yamauchi, T. Yamashita, H.H. Fujiyoshi. "Boosted random forest." IEICE Transactions on Information and Systems, vol. 98, 9, 2015, pp. 1630-1636.
10. [10] URL: https://web.stanford.edu/class/archive/cs/cs221/cs221.1186/lectu res/learning3.pdf (Access: Jan 5, 2023).
11. [11] G. Leshem, Y.A. Ritov. "Traffic flow prediction using Adaboost algorithm with random forests as a weak learner." International Journal of Mathematical and Computational Sciences, vol. 1, 1, 2007, pp. 1-6.
12. [12] H. Abdulsalam. "Streaming random forests." Ph.D. Thesis, Queen’s University, Canada, July 2008.