The Study And Simulation Of Pedestrian Evacuation With Fog Based On Agent

Since the beginning of the twenty-first century, several major public security events caused by congestion or sudden catastrophe have happened continuously in worldwide. These events have caused a large number of casualties and property losses, and brought unnecessary damages to human society and economics. In many cases of evacuation, evacuation in fire is one of the most common and dangerous. Especially in crowded places, the fire disasters include not only the life and property losses caused by the fire itself, but also including damages caused by stampede due to the panic of people in fire as an addition. With the improvement in the level of development of economic construction and personnel-intensive, the occurrence of fire accidents in China is getting worse. In order to study the laws of pedestrian movement better and develop evacuation strategies, the researchers have constructed series models in a view of simulation for this problem. They expect the computer such a modern powerful computational tool can help them to provide a new way to study pedestrian evacuation problems.At present, the research of fire evacuation at home and abroad is mainly focused on the affection of "fire" much more than the affection of "smoke". In fact, the major death in fire disaster is not caused by the fire, but caused by the smoke of fire. To the pedestrian evacuation with fog, the researchers in the internal generally take a separating method. On one hand, they focus on the study of smoke generation and diffusion laws. On the other hand, they focus on the study of the pedestrian evacuation strategy and process in this dangerous environment. This method of course benefits to split a complex problem into some simple ones in order to study them individually, but obviously it also lacks a whole, real-time considerations to the pedestrian evacuation with fog such a consistent problem. Just from this aspect, the contents of this article can be developed rapidly.In order to simulate spread of smoke, the paper discussed a variety of common gas dispersion modeling methods. By comparison, this paper considers that the use of microscopic simulation can well reproduce the actual movement of the gas diffusion. The method is based on the kinetic theory of gas molecules, which assumes that the gas molecules do irregular thermal motion, and obtains the whole motion of the diffusion of gas by simulating a large number of random motion of molecules. So, it has characteristics that the model is simple and intuitive, and the process is true and reliable.In short, this paper take an agent-based modeling technique, which studies the problem of pedestrian evacuation in the room with the spread of smoke. This paper discussed in detail such problems:the law of the spread of smoke, the effect to pedestrian evacuation by smoke generating location and smoke diffusing rate, and the meaning of firemen to pedestrian evacuation with fog. Through the simulation, this paper finds that the closer smoke generating location distance to door, and the faster is smoke diffusing rate compared to pedestrian moving rate, the lower is pedestrian evacuation efficiency, and the less pedestrians survive. But after leading the firemen in, by firemen aiding pedestrian to evacuate and escape, the number of survival pedestrians became larger while the number of the firemen increased. The image of data shows the characteristics of a concave function. In this article, the firemen actually play a role on guiding path, and aiding to continue life. In real life, these purposes can also be achieved by marking the direction of exports, increasing the gas masks and other rescue facilities. This paper considered that it has important meanings for pedestrians’ successful evacuation in a fog environment when fire to take these measures.The reason of taking agent-based modeling approach in this article is that this paper considers it is an advanced technology and concept. It not only has many good properties belonged to object such as encapsulation, inheritance, and polymorphism, but also has several more advanced nature like initiative and sociality. This article considers that micro-simulation modeling of pedestrian flow will be developed finer and larger in the future. So the ability of intelligent and distributed computing is required, and the agent technology is just fit for these requirements. Furthermore, the article also envisaged a unified data modeling of pedestrian flow, and considered that this unified data modeling is more conducive to the communication among pedestrian flow researchers, as well as the standardization of the model.