| Nowadays it is a world of information technology development which has brought the geometric growth .This process promoted the study of human behavior. In the past five yeas, the study of human behavior has developed into a single subject and many socio-economic understanding of complex systems is based on scientific analysis of human behavior .On the present day, a experimental study on a variety of statistical characteristics of human behavior found that the phenomenon of human behavior have non-Poisson properties, and its interval distribution is similar to power-low distribution. The fat tail is a popular characteristic which exists in the time interval between events of human behaviors. This discovery marks research into human behavior to a new area. The important research area of dynamics of human behavior is about the spatial characteristics of human behavior study, this thesis is to study the movement path of human behavior. For the study, the establishment of an appropriate model is a top priority.In this study, the first consideration is that the memory effect and distance effect influence on human behavior, based on the classical random walk theory, a dynamic model of a human space. And give it a numerical simulation and analytical study, according to the results of model simulation of the model. The model is modified. Put the traffic level theory into the use of the model. The analysis of Model simulation results uses likelihood estimation the exponential fit and uses the Kolmogorov-Smirnov test used method to test the authenticity of the distribution curve. The study shows that the model can get index between 0 and 3 for the power-law form of the interval distribution which is consistent with some empirical results. This result suggests a spatial effect and memory effect is characteristic of human behavior, one of the causes of non-Poisson. But when it is used in the theory of the traffic the results of modified model is more accuracy. The importance of location may also be a characteristic of human behavior, the reasons for non-Poisson properties. |