Abstract
Problem Statement (Relevance): This article examines contemporary structural models [1] designed for predicting the M25 and М10 coke quality indices, which are used for theoretical research and in practical application by Russia’s coke producers. For the studied models, the model accuracy indicators were brought to a single format and the feagures were compared with the specification given in GOST 5953-93 “+20 mm coke. Physical strength determination” (ISO 556-80). Objectives: The objectives include a comparison and an adequacy analysis of the mathematical models designed for predicting the M25 and M10 coke quality indices based on the models’ predictive accuracy. Methods Applied: The authors carried out a statistical analysis of the confidence intervals for the prediction error of mathematical models designed for predicting the M25 and M10 coke quality indices. The authors looked at the length of the confidence interval at various confidence probabilities before drawing conclusions on the models’ predictive accuracy. Originality: This work provides an opportunity to verify the mathematical models for predicting the M25 and M10 coke quality indices based on the fact that the predictive accuracy of the model should not exceed the allowable discrepancy between two parallels when analyzing the coke quality per GOST 5953. Findings: The article provides a comparison between different structural models in terms of their predictive accuracy. This comparison helped analyze the adequacy of the mathematical models. Practical Relevance: The results obtained can be used for theoretical research and in practical application by Russia’s coke producers.
Keywords
Coke, coke quality indices, mathematical model, classification on the basis of internal structure and algorithm, structural model.
1. Smirnov A.N., Petukhov V.N., Alekseev D.I. Looking at how mathematical models are built for predicting the M25 and M10 coke quality indices for the purposes of classification and developing the concept of a hybrid model. Koks i khimiya [Coke and chemistry], 2015, no. 5, pp. 13–18. (In Russ.)
2. Stankevich A.S., Zolotukhin Yu.A., Kalinina G.I., Aleshin V.I., Chemerko V.A. A relationship between the physical strength of coke, the chemical and petrographic composition of burden material made with the Kuzbass coal and the coking process. Koks i khimiya [Coke and chemistry], 1981, no. 2, pp. 27–31. (In Russ.)
3. Stankevich A.S., Myuller I.P., Lel'chuk V.I. Coal blending for coking and linear programming for predicting the coal quality. Koks i khimiya [Coke and chemistry], 1981, no. 11, pp. 4–8. (In Russ.)
4. Stankevich A.S. Charge calculation and predicting the quality of coke made with coal from the Asian basins based on its terographic characteristics. Koks i khimiya [Coke and chemistry], 1983, no. 9, pp. 11–16. (In Russ.)
5. Stankevich A.S., Tregub V.V., Aleshin V.I., Fisich M.V., Gribovich A.V., Tsinker L.Z. Predicting the coke quality based on the Unified Industrial Coal Classification. Koks i khimiya [Coke and chemistry], 1990, no. 12, pp. 36–39. (In Russ.)
6. Stankevich A.S., Stepanov Yu.V., Gilyazetdinov R.R., Popova N.K. Predicting the physical strength of coke on the basis of the chemical and petrographic composition of coal burden material. Koks i khimiya [Coke and chemistry], 2005, no. 12, pp. 14–21. (In Russ.)
7. Stankevich A.S., Stankevich V.S. Determining the value of coals used for coke production. Koks i khimiya [Coke and chemistry], 2011, no. 6, pp. 2–10. (In Russ.)
8. Stankevich A.S., Stankevich V.S. Methods for determining the coking properties and the value of coal and coal mixtures. Koks i khimiya [Coke and chemistry], 2012, no. 1, pp. 4–12. (In Russ.)
9. Stankevich A.S., Bazegskyi A.E. Optimizing the quality of the EVRAZ ZSMK coke with regard to the properties of the available coal. Koks i khimiya [Coke and chemistry], 2013, no. 10, pp. 14–21. (In Russ.)
10. Ammosov I.I., Eremin I.V., Sukhenko S.I., Oshurkova L.S. Calculation of coking charges based on the petrographic properties of coals. Koks i khimiya [Coke and chemistry], 1957, no 12, pp. 9–12. (In Russ.)
11. Eremin I.V., Gagarin S.G. Calculation of coking charges on the basis of a petrographic model. Koks i khimiya [Coke and chemistry], 1992, no. 12, pp. 9–15. (In Russ.)
12. Gagarin S.G., Eremin I.V. Computer monitoring of coke strength on the basis of a petrographic coal charge calculation model. Koks i khimiya [Coke and chemistry], 1995, no. 2, pp. 10–15. (In Russ.)
13. Gagarin S.G. Coke strength petrographic prediction model assessment – A case study of Mongolian coals. Koks i khimiya [Coke and chemistry], 2011, no. 4, pp. 21–26. (In Russ.)
14. Venttsel' E.S. Teoriya veroyatnostey [Probability theory]. 10th edition. Moscow: Akademiya Publishing Centre, 2005, 576 p. (In Russ.)
15. Kiselev B.P., Stukov M.I., Olshanetskyi L.G., Leushin B.A., Tyagunov L.I. A method for predicting the coke strength and calculating the coal charge mixtures. 1. Verification and assessment of the prediction methods proposed. Koks i khimiya [Coke and chemistry], 1990, no. 3, pp. 34–37. (In Russ.)