Розробка методу адаптивного визначення швидкості руху людини при моделюванні поведінки натовпу в екстремальних ситуаціях

The subject of this study is the dynamics of human flow at extreme events, which are modeled using a computer simulation model. The task addressed relates to the insufficient accuracy in determining human speed with existing approaches when modeling crowd behavior at extreme events. Specifically, the desired walking speed of a person is set fixed or discretely, which can lead to significant errors as it does not correspond to real-world conditions. Models based on pre-collected data compromise accuracy under different conditions. A method is proposed to adaptively determine human movement speed at extreme events, which takes into account individual spatial constraints and the narrowing of the effective field of view under stress. Simulation modeling has shown that the method devised significantly improves the accuracy of the models. The average modeling error decreased from 28.05 % to 12.06 % for a circular profile of human projection, and from 31.5 % to 6.09 % for an elliptical profile. The results are explained by the individual consideration of local crowd density, realistic narrowing of the field of view within the range of 30° to 0.5°, and corresponding adaptive adjustment of the desired speed. A feature of the devised method is its universality as it does not depend on a specific scenario or pre-collected empirical data. The method is based on general patterns of human interaction with the environment and is therefore suitable for use even in cases where field studies are impossible or difficult. Provided that two-dimensional models are used, the proposed method could be applied to simulate crowd behavior in automated crowd management systems, software packages for assessing the safety of mass events, and designing evacuation routes