A Pedestrian Walking Behavior Prediction Method Based on Markov State Transition

Abstract

The invention discloses a method for predicting a pedestrian walking behavior based on the Markov state jump. The method comprises the steps that S1, movement features of a crowd at a crowded state are investigated, and statistical analysis is performed on the acquired information to obtain a density-speed relation and a density-flow relation of the crowd; S2, an information transfer based direction selection model is built according to influences imposed on the path search and selection process of a crowd movement behavior by factors such as view field condition of a pedestrian in the movingprocess, the building scale and information acquisition; S3, discretization is performed according to the movement speed of the crowd in a relation diagram, and the movement speed of the pedestrian isselected by using an idea of Markov state jump; and S4, the pedestrian reduces the speed in advance when encountering an obstacle and keeps a certain distance with the obstacle to avoid contact and collision. The movement behavior of the dense crowd can be accurately predicted according to the movement of the pedestrian under the rule, thereby providing a reference for correctly dispersing the crowd. Meanwhile, the method can improve the space structure design of large public places and has great significance for improving the crowd evacuating efficiency and ensuring the life safety.

Description

technical field [0001] The invention relates to the field of urban pedestrian traffic safety, and more particularly, to a pedestrian walking behavior prediction method based on Markov state transitions. Background technique [0002] In recent years, with the development of economy and culture, the scale of the city has expanded rapidly, and the urban population has grown rapidly. The traffic problem has become one of the most concerned issues. Pedestrian movement is an important part of road traffic, which has a great impact on traffic flow and traffic management. At present, the research on pedestrians in road traffic has been paid more and more attention. Crowd gathering activities are increasingly appearing in various public places. When the number of people in the place increases and the density is high, pedestrians will interweave and squeeze each other, and a small disturbance will make the crowd enter an unstable state. Effective control and management can easily cau...

AI Technical Summary

Problems solved by technology

Traditional research methods include cellular automata model and social force model. The cellular automaton model statically processes space and homogenizes the rules of a single pedestrian. The algorithmic complexity of the social force model increases exponentially with the increase in the number of people. and cannot reflect the randomness of a single pedestrian's behavior

Images