The Study And Implementation Of The Simulation System Of Crowd Evacuation

As the concentration of wealth in society, the increasing density of urban population, a variety of unexpected events occur frequently, for a gathering place when hold a large event, the evacuation of people must be fully prepared to conduct evacuation, by the assessment of simulation of virtual people of emergency evacuation program is of great practical significance and value in use.For the problem of imperfect mechanism of decision-making and rules of the behavior which base on agent simulation, this paper presents the method of base on principles of fuzzy logic which meets the characteristics of thinking of brain and reasoning system rule set of human, making the agent can independently produced reasonable goal, and then planning the shortest path to the target.By building the multi-level navigation map simplifies the virtual scene graph.Based on studies of the Dijkstra algorithm and A* algorithm, provide the optimal path planning methods.For the problem about the low performance of the algorithm when in the large scene,we proposed two ways to improve the path planning performance.Agent Simulation System in all the movements are the combination or superposition generated of some basic manipulation. By analyzing the characteristics of autonomous agent,we design various manipulation in a virtual environment, we design group behavior of the agent at the same time.Based on the analysis and study of the simulation of crowd behavior, we build a distributed simulation system architecture which can be extended and developed an emergency simulation system to disperse the crowd. The results of simulation show:the results are close to reality and the key methods and techniques are effective.