Clustering of fires at fuel and energy complex facilities using retrospective statistical data to identify fires ranks

Introduction. Such a definition as “fire rank” is used in the practice of fire respond and elimination management for quite a long time and has several meanings, the main of which are two: to show in detail the degree of complexity of the fire based on the fire rank, to allocate an adequate number of different types of resources for its successful elimination. The chain of elements: fire — degree of complexity — necessary resources, has not yet been embodied in effective methods and regulatory documents, which gives relevance to this area of research.

Goals and objectives. The purpose of the paper is to build a technology for constructing models of fire ranks based on the use of cluster analysis. The tasks include exploratory data analysis and construction of a stochastic model of fire ranks.

Methods. The study used methods of mathematical statistics and unsupervised machine learning in the variant of cluster analysis.

Results and discussion. The possibility of determining a grid of fire ranks and the corresponding number of allocated vehicles based on processing by cluster analysis methods of a selection of retrospective data is shown. It is shown that the obtained results are very close to the norms of allocation of vehicles for fires in Moscow. The concept of stochastic fire ranks as a more informative model for resource allocation is put forward.

Conclusions. The presented results of solving the problem of identifying a grid of fire ranks using a selection of retrospective data make it possible not only to determine the amount of resources required to eliminate a particular fire, but also make it possible to build adaptive stochastic models of fire ranks that are adequate to the region, industry and other subsets of fire objects.

1. Topolskiy N.G., Vilisov V.Ya. Methods, models and algorithms in security systems: machine learning, robotics, insurance, risks, control. Moscow, RIOR Publ., 2021; 475. DOI: 10.29039/02072-2 (rus).

2. Topolskiy N.G., Prus Yu.V., Klimovtsov V.M. Determining the rank of fire at a facility using state diagrams. Materials of the 13th international conference “Security Systems” SB-2004. Moscow, State Fire Service Academy EMERCOM of Russia, 2004; 297-299. (rus).

3. Topolskiy N.G., Butuzov S.Yu., Vilisov V.Ya. Information and analytical models of management support in the elimi­nation of fires and emergencies. Moscow, State Fire Service Academy EMERCOM of Russia, 2021; 216. (rus).

4. Vilisov V.Ya., Khabibulin R.Sh. Application of experimental planning methods in expert procedures for machine learning of a resource distribution model during fire extinguishing. Technology of Technosphere Safety. 2023; 1(99):44-63. DOI: 10.25257/TTS.2023.1.99.44-63 (rus).

5. Khabibulin R.Sh. Cluster analysis in the field of prevention and response to fires and emergencies. Technology of Technosphere Safety, 2022; 3(97):202-214. DOI: 10.25257/TTS.2022.3.97.202-214 (rus).

6. Vilisov V.Ya., Khabibulin R.Sh. Statistical analysis and modeling data on eliminating fires at fuel and energy enterprises. Fire and emergencies: prevention, elimination. 2023; 3:63-74. DOI: 10.25257/FE.2023.3.63-74 (rus).

7. Stroh R., Bect J., Demeyer S., Fischer N., Vazquez E. Gaussian process modeling for stochastic multi-fidelity simu­lators, with application to fire safety. 2016. DOI: 10.48550/arXiv.1605.02561

8. Lee H.-R., Lee T. Multi-agent reinforcement learning algorithm to solve a partially-observable multi-agent problem in disaster response. European Journal of Operational Research. 2021; 291(1):296-308. DOI: 10.1016/j.ejor.2020.09.018

9. Bedilo M.V., Butuzov S.Yu., Prus Yu.V., Ryzhenko A.A., Chursin R.G. Model of adaptive management of emergency services of the RSChS in emergency situations at the interregional and federal levels. Technology of Technosphere Safety. 2017; 1(71). URL: http://agps-2006.narod.ru/ttb/2017-3/44-01-17.ttb.pdf (rus).

10. Hamke E.E., Jordan R., Ramon-Martinez M. Breath activity detection algorithm. 2016. DOI: 10.48550/arXiv.1602.07767

11. Zhang X., Mahadevan S. Bayesian neural networks for flight trajectory prediction and safety assessment. Decision Support Systems. 2020; 131. DOI: 10.1016/j.dss.2020.11324612.

12. Vlasov K.S., Poroshin A.A. Study of regional characteristics of the parameters of large fires. Technology of Technosphere Safety. 2022; 2(96):82-91. DOI: 10.25257/TTS.2022.2.96.82-91 (rus).

13. Nikolenko S., Kadurin A., Arkhangelskaya E. Deep learning. Dive into the world of neural networks. St. Petersburg, Piter publ., 2018; 480. (rus).

14. Khaykin S. Neural networks: a complete course. Moscow, Williams Publ., 2006; 1104. (rus).

15. Pedregosa F., Varoquaux G., Gramfort A., Michel V. et al. SciKit-Learn. Machine learning in Python. Journal of Machine Learning Research. 2011; 12:2825-2830.

16. Fertier A., Barthe-Delanoë A.-M., Montarnal A., Truptil S., Bénaben F. A new emergency decision support system: the automatic interpretation and contextualisation of events to model a crisis situation in real-time. Decision Support Systems. 2020; 133. DOI: 10.1016/j.dss.2020.113260

17. Cavdur F., Sebatli A. A decision support tool for allocating temporary-disaster-response facilities. Decision Support Systems. 2019; 127. DOI: 10.1016/j.dss.2019.113145

18. Dyuran B., Odell P. Cluster analysis. Moscow, Statistics Publ., 1977; 127. (rus).

19. Pranov B.M. Issues of application of cluster analysis in fire statistics. Technology of Technosphere Safety. 2021; 4(94):117-124. DOI: 10.25257/TTS.2021.4.94.117-124 (rus).

20. Fires and fire safety in 2019. Statistics of fires and their consequences : statistical collection. D.M. Gordienko (rus. ed.). Moscow, VNIIPO MChS Rossii, 2020; 80. (rus).

21. Zhang T., Wang Z., Wong H.Y., Tam W.Ch., Huang X., Xiao F. Real-time forecast of compartment fire and flashover based on deep learning. Fire Safety Journal. 2022; 130:103579. DOI: 10.1016/j.firesaf.2022.103579

22. Nguyen H.T., Abu-Zidan Y., Zhang G., Nguyen K.T.Q. Machine learning-based surrogate model for calibrating fire source properties in FDS models of facade fire tests. Fire Safety Journal. 2022; 130:103591. DOI: 10.1016/j.firesaf.2022.103591

23. Gulyamova G.M. About methods of decrease in fire hazard at gas stations. International Academy Journal Web of Scholar. 2019; 1(1):8-10. DOI: 10.31435/rsglobal_wos/31012019/6306