Modern approaches to studying human behavior in a critical situation: integration of 3D agents into virtual reality

In modern conditions, modeling human behavior in critical situations, such as fire, is becomingparticularly relevant to improve the safety and effectiveness of evacuation processes. The developmentof virtual reality (VR) technologies opens up new opportunities for research and analysis of suchsituations. The aim of the research is to develop a 3D simulation model of agents that allows analyzingtheir behavior in critical situations and predicting evacuation routes. The main areas of work includebuilding a mathematical model of agent movement, implementing evacuation algorithms in a virtualenvironment, and integrating the results into game engines for visualization and analysis. The scientificsignificance of the research lies in the development of algorithms that take into account the peculiaritiesof human interaction in complex architectural conditions and allow analyzing the behavior of agents invarious scenarios. The practical significance of the work lies in the fact that the created model can be used for staff training, testing evacuation plans and improving the design of buildings in terms of safety.The research methodology includes mathematical modeling of agent behavior, the use of optimizationalgorithms to calculate evacuation routes, as well as the integration of the model into a virtualenvironment using the Unreal Engine (UE). As a result of the research, an approach to modeling is proposed that allows predicting the movement of people in a fire, evaluating the effectiveness of evacuation routes and analyzing the influence of various factors on the evacuation process. The resultsobtained can be used in safety research, the development of simulators for emergency services and the improvement of architectural solutions.