Dynamic Modeling And Complicated Characteristics Of Public Opinion Propagation And Pedestrian Traffic Based On Game Psychology

With the rapid development of China’s economy, the material and culture demand of people is growing, walking street, public square, subway and other public places have become an important part of the broad masses of the people in daily life. However, these places are relatively dense, more mobile and much easier to induce pedestrian jams, and lead to the occurrence of all kinds of accidents. Therefore, how to scientific management and effective control pedestrian congestion is a serious problem need be urgently solved. Because the pedestrian system is a typical self-driven system and far away from equilibrium state, pedestrians not only have strong interaction among themselves, but also have many significant characteristics, such as independence, flexibility and randomness etc. Therefore, how to combine statistical physics, nonlinear dynamics, traffic engineering and other multi-disciplinary theories and research methods, and establish reasonable pedestrian flow models to explore the macroscopic characteristics and evolution law of the pedestrian movement, and disclose formation mechanism of pedestrian congestion, which are great significance for us to guide the pedestrian traffic planning, management, control and avoiding the jam of pedestrian flow. In this dissertation, based on the cellular automata models and lattice gas models, several improved dynamic models are proposed to investigate public opinion, unidirectional, bidirectional, four-way and curve pedestrian flow by considering the psychological factors, and introducing game theory and learning strategy. And the macroscopic characteristics of pedestrian flow and the formation mechanism of pedestrian jams are studied via theoretical analysis and numerical simulation. In addition, we put forward some valuable suggestions for the planning of pedestrian facilities in channel, walking street and curve region, and management and controls of pedestrian flow from a qualitative point of view.The main contents of the dissertation are as follows.Ⅰ. A modified cellular automaton model is proposed to simulate opinion propagation and evolution by considering the difference between individuals, and introducing an obstinate strength parameter to reflect the preferential original attitude habits of people. The obstinate strength, occupancy of the obstinate people, system size and propagation mechanism of public opinion are investigated.To explore the influences of obstinate people on opinion dynamics, a modified cellular automaton model is proposed to simulate opinion propagation and evolution by introducing an obstinate strength parameter representing the people preferential original attitude habits. The influences of obstinate strength, obstinate people’s occupancy and system size are investigated via numerical simulation. Furthermore, the typical space-time patterns are also discussed in detail. The simulation results show that only the larger obstinate strength has an important effect on final-state approval rate. The larger occupancy for obstinate people, the easier achieve common attitude. The final state of the system is completely independent of the system size. The results above could regulation and guidance for the mass media to provide a useful reference.Ⅱ. Aiming at the location conflict of pedestrian movement, and taking into account the psychological changes of pedestrians, and introducing game behaviors and learning strategies, unidirectional and bidirectional lattice gas model are proposed to simulate pedestrian flow, respectively. The influences of the different game behaviors, the length and position of central barrier on the pedestrian flow are studied.Considering the psychological changes of pedestrian, the game behaviors and learning strategy are introduced to describe decision-making behavior of pedestrian, when face to location conflicts. And the corresponding single-directional and bi-directional pedestrian flow lattice gas models are established, respectively. The influences of different game behavior, the length and position layout of central barrier on pedestrian flow are investigated via numerical simulation. Furthermore, the efficiency of pedestrian movement also is discussed in detail. The simulation results show that central barrier has almost no effect on the movement efficiency of the unidirectional pedestrian flow. However, the central barrier has an obvious influence on the bidirectional pedestrian flow. In addition, two central barriers are more effective to improve pedestrian total movement efficiency than that of one central barrier.Ⅲ. Based on Tajima model, an extended lattice gas model is proposed to simulate bidirectional pedestrian flow by considering the visual field effect, and the generation mechanism of macroscopic phenomena induced by pedestrian’s visual field are investigated.Through considering the game behaviors and learning strategies, an extended lattice gas model is established to describe pedestrian counter flow. The effect of visual field is explored, and the characteristics of space-time dynamics also are discussed. The simulation results show that (1) the model can better reproduce some essential features of pedestrian counter flow, such as the lane formation and segregation effect. (2) at low density, pedestrian should move toward the visual field region of small numbers of countermove pedestrians, which can avoid collision among pedestrians and helpful improve movement efficiency of pedestrian flow. On the contrary, at high density, if pedestrian choice following movement, it can effectively reduce the occurrence of pedestrian jams.Ⅳ. Based on the Muramatsu model, an improved lattice gas model is proposed to simulate four-way pedestrian flow (e.g., crossroads etc.) by considering game behavior, and the influence of game parameters, system size and other factors on the macroscopic characteristics of pedestrian flow are explored.Based on the Muramatsu model, an extended lattice gas model is established to simulate four-way pedestrian flow. The influences of game parameters, system size and other factors on pedestrian flow are studied via numerical simulation. The results show that:the game behaviors of pedestrians are not always advantageous to the pedestrian’s movement. When the density is low, the game behaviors have almost no effect on the evacuation time of pedestrian. At high density, the blindness game behavior will lead to pedestrian jams, it is helpful to improve the evacuation efficiency of pedestrians by increasing the game probabilities of two or three people. At the same density, the larger system size can improve the efficiency of pedestrian movement.Ⅴ. A novel lattice gas model is proposed to simulate pedestrian flow in curve channel by considering the right movement tendencies of pedestrian, and the influence of with or without the central barrier and other factors on pedestrian flow are studied.Considering pedestrian’s walking habits (e.g., right movement tendencies, etc.), an extended lattice gas model is established to simulate pedestrian flow in curve channel. The effect of right movement tendencies, central barrier and other factors are investigated via numerical simulation. The results show that:in the absence of the central barrier, increasing the strength of the right movement tendencies can improve the efficiency of pedestrian movement; the existence of the central barrier can reduce the occurrence of pedestrian jams to some extent. In addition, the reasonable length of central barrier is helpful to improve pedestrian movement, but the long central barrier will lead to pedestrian congestion. The final chapter of this dissertation is devoted to a summary of the thesis and a prospect of further study of the pedestrian flow.