Intelligent analysis of smoke formation processes in fires: new approaches based on AI technologies

Abstract

Objective. The aim of the study is to examine approaches to creating an intelligent system, the implementation of which can affect the quality of decisions made by the fire extinguishing manager (FEM) in the process of managing personnel during fire extinguishing. Method. The work uses methods of statistical analysis, control theory, data processing methods and the implementation of production rules for indexing, hashing and clustering of the received information for its further transformation into a knowledge base. Result. The paper presents statistical data on injuries and deaths of fire and rescue unit personnel from thermal consequences caused by the movement of gaseous masses at capital construction sites. The research cycle on defining and applying artificial intelligence methods and algorithms in the process of managing combat operations at the call site is continued. The inadequacy of studies of the smoke-forming ability of building materials affecting the personnel of the Russian Emergencies Ministry at the call site is determined. Conclusion. An algorithm for creating a database of images of smoke generated during combustion of various substances and materials that constitute the main combustible load in the premises of objects of different functional purposes is formulated, and a classifier of smoke images depending on the fire load is proposed. An example of a description of the structure of the database being formed based on existing simulation models and software products is given. A model for forming a knowledge base of the RTP for smoke analysis depending on the type of combustible load is presented.