D.S. Korolev1, Ph.D. of Engineering Sciences; A.N. Baturo2, Ph.D. of Engineering Sciences
1Voronezh state technical university (high school)
2Siberian Fire and Rescue Academy of Firefighting Service of EMERCOM of Russia
Artificial intelligence technologies are actively introduced in various industries of the country, including in the field of fire safety of oil and gas complexes. Today, there are more than 110 million synthesized substances, and their fire hazard properties have been partially studied and only for a few thousand compounds. This creates difficulties in calculating the intensity of the thermal radiation of the flow for fire strait limit ketones. To solve the problem, we applied a method for predicting the fire hazard properties of oil products based on molecular descriptors and artificial neural networks. The specific mass burnup rate and the average surface density of the thermal radiation of the flame for a sample of substances were predicted, a comparative analysis was carried out. The relative error is determined, not exceeding 5%. Thus, using a digital approach to solving the problem of the lack of information on the properties of substances, we have expanded the existing database, which can be used to develop a fire safety system.
Открыть » // Ссылка для цитирования этой статьи