Fuzzy logic and machine learning methods applied to the analysis of industrial power consumption under the condition of uncertainty

Abstract

Introduction. Recently, the fuzzy logic method has been widely implemented in solving various problems of economic research, including theoretical analysis of the economy and its resource dependence, the study of innovative processes in a resource-type economy.

Purpose. The purpose of the research is to analyze the dependence of industrial power consumption from various social economic factors with the fuzzy modeling method. This method is particularly well suited for modeling ill-defined systems with the significant uncertainty about the nature and range of key input variables and the underlying relationships between them. This system could be illustrated by the economy of modern Russia at the time of sanctions imposed by unfriendly states.

Materials and methods. The work refers to fuzzy modeling and machine learning methods. A random forest algorithm was used to select predictors and for comparative analysis.

Results. The results of fuzzy modeling were compared with the results obtained by modeling the analyzed relationship with multiple regression, and with the results obtained by applying the random forest method with regression decision trees to the data under study. Fuzzy logic-based modeling of the above-described dependence in the context of uncertainty is shown to be more adequate compared to regression-based modeling (including the random forest method).

Conclusion. The proposed fuzzy system (fuzzy inference system) can be used to study the influence of changes in any input factor or their combination on changes in industrial power consumption. The fuzzy system could reveal how much various production locations could change industrial electricity consumption or analyze the feasibility of a location in terms of access to labor resources. It is also possible to study how much the number of employees associated with the outflow of labor resources could change industrial electricity consumption.

Keywords: fuzzy logic, neuro-fuzzy inference, power consumption, random forest method, machine learning, multiple regression

For citation

Serkov L. A. Fuzzy logic and machine learning methods applied to the analysis of industrial power consumption under the condition of uncertainty. Perm University Herald. Economy, 2024, vol. 19, no. 1, pp. 52–68. DOI 10.17072/1994-9960-2024-1-52-68. EDN BTNLUG.

REFERENCES

1. Zhaoying O. Construction and Application of Economic Management Fuzzy Decision Model Based on Fuzzy Relevance Method. Journal of Mathematics, 2022, vol. 2022, Article ID 9878815, 11 p. DOI 10.1155/2022/9878815
2. Jang J.-S. R. ANFIS: adaptive-network-based fuzzy inference system. IEEE Transactions on Systems, Man and Cybernetics, 1993, vol. 23, iss. 3, pp. 665–685. DOI 10.1109/21.256541
3. Abraham A. Adaptation of fuzzy inference system using neural learning. Nedjah N., Macedo Mourelle L. (eds) Fuzzy Systems Engineering. Studies in Fuzziness and Soft Computing, 2005, vol. 181, Berlin, Springer, 2005, pp. 53–83. DOI 10.1007/11339366_3
4. Giles D. E. A. Output convergence and international trade: Time-series and fuzzy clustering evidence for New Zealand and her trading partners, 1950 –1992.  Journal of International Trade and Economic Development, 2005, vol. 14, iss. 1, pp. 93–114. DOI 10.1080/0963819042000333261
5. Lebedeva М. Fuzzy logic in economics – the formation of a new direction. Idei i idealy = Ideas and Ideals, 2019, vol. 11, no. 1-1, pp. 197–212. (In Russ.). DOI 10.17212/2075-0862-2019-11.1.1-197-212. EDN ZCWNFJ
6. Giles D. E. A., Draeseke R. Econometric modeling based on pattern recognition via the fuzzy C-means clustering algorithm. Working Paper EWP0101. Department of Economics, University of Victoria. 2001. 50 p. Available at: https://www.uvic.ca/socialsciences/economics/_assets/docs/econometrics/ewp0101.pdf (access date 20.01.2023).
7. Takagi T., Sugeno M. Fuzzy identification of systems and its applications to modeling and control. IEEE Transactions on Systems, Man, and Cybernetics, 1985, vol. SMC-15, no. 1, pp. 116–132. DOI 10.1109/TSMC.1985.6313399
8. Xiao Y., Liu J. J., Hu Y., Wang Y., Lai K. K., Wang S. A neuro-fuzzy combination model based on singular spectrum analysis for air transport demand forecasting. Journal of Air Transport Management, 2014, vol. 39, pp. 1–11. DOI 10.1016/j.jairtraman.2014.03.004
9. Kumar S., Singh R., Manish K., Ashish K. Fuzzy logic based model to calculate the economic level of any country. International Journal of Basic Sciences and Applied Computing (IJBSAC), 2016, vol. 1, iss. 12, pp. 6–12. Available at: https://www.ijbsac.org/wp-content/uploads/papers/v1i12/L00570311216.pdf (access date 20.01.2023).
10. Petrov M. B., Serkov L. A., Kozhov K. B. Assessment of energy supply to the Yamal peninsula based on fuzzy multicriteria analysis. Ekonomika regiona = Economy of Regions, 2022, vol. 18, no. 4, pp. 1209–1222. (In Russ.). DOI 10.17059/ekon.reg.2022-4-17. EDN BAVQIV
11. Shipley M., Johnson M., Pointer L., Yankov N. A fuzzy attractiveness of market entry (FAME) model for market selection decisions. Journal of the Operational Research Society, 2013, vol. 64, iss. 4, pp. 597–610. DOI 10.1057/jors.2012.59
12. Neocleous C. K., Schizas C. N., Papaioannou M. C. Fuzzy cognitive maps in estimating the repercussions of oil/gas exploration on politico-economic issues in Cyprus. 2011 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE 2011). Taipei, Taiwan, 2011, pp. 1119–1126. DOI 10.1109/FUZZY.2011.6007655
13. Cummins J. D., Derring R. A. Fuzzy trends in property-liability insurance claim costs. The Journal of Risk and Insurance, 1993, vol. 60, no. 3, pp. 429–465. DOI 10.2307/253037
14. Tselykh A. N., Tselykh L. A., Prichina O. S. Metody nechetkoi logiki v upravlenii proizvodstvennymi protsessami. Izvestiya YuFU. Tekhnicheskie nauki = Izvestiya SFedU. Engineering Sciences, 2014, no. 1 (150), pp. 111–119. (In Russ.). EDN RTZMYH
15. Alaqeel T., Suryanarayanan S. A fuzzy analytic hierarchy process algorithm to prioritize smart grid technologies for the Saudi electricity infrastructure. Sustainable Energy, Grids and Networks, 2018, vol. 13, pp. 122–133. DOI 10.1016/j.segan.2017.12.010
16. Mahjouri M., Ishak M., Torabian A., Manaf L., Halimoon N., Ghoddusi J. Optimal selection of iron and steel wastewater treatment technology using integrated multi-criteria decision-making techniques and fuzzy logic. Process Safety and Environmental Protection, 2017, Vol. 107, pp. 54–68. DOI 10.1016/j.psep.2017.01.016
17. Zimmer K., Fröhling M., Breun P., Schultmann F. Assessing social risks of global supply chains: A quantitative analytical approach and its application to supplier selection in the German automotive industry. Journal of Cleaner Production, 2017, vol. 149, pp. 96–109. DOI 10.1016/j.jclepro.2017.02.041
18. Bellman R., Zade L. Prinyatie reshenii v rasplyvchatykh usloviyakh. Voprosy analiza i protsedury prinyatiya reshenii = Issues of Analysis and Decision Making Process, Moscow, 1976, pp. 172–215. (In Russ.).
19. Leonenkov A. Nechetkoe modelirovanie v srede Matlab i fuzzyTECH. St Petersburg, 2005. 719 p. (In Russ.).
20. Breiman L. Random forest. Machine Learning, 2001, vol. 45, pp. 5–32. DOI 10.1023/A:1010933404324
21. Svetnik V., Liaw A., Tong C., Culberson J. C., Sheridan R. P., Feuston B. P. Random forest: A classification and regression tool for compound classification and QSAR modeling. Journal of Chemical Information and Computer Sciences, 2003, vol. 43, iss. 6, pp. 1947–1958. DOI 10.1021/ci034160g
22. Petrov M. B., Serkov L. A., Kozhov K. B. Analysis of the spatial features of regional power consumption in the Russian Federation. Applied Econometrics, 2021, no. 1 (61), pp. 5–27. (In Russ.). DOI 10.22394/1993-7601-2021-61-5-27. EDN PGLLIM
23. Chiu S. Fuzzy model identification based on cluster estimation. Journal of Intelligent and Fuzzy Systems, 1994, vol. 2, iss. 3, pp. 267–278. DOI 10.3233/IFS-1994-2306

Show full text