基于马尔科夫链的移动轨迹预测模型研究

Personal activities seemingly out of order as molecular motion, but there are potential models. Relevant scholars pointed out that the population mobility has a non random and periodic, people always appear in a particular place, no matter how the diversity of activities, activities of the crowd always follow the simple reproduction mode, predictable height. For people with mobility, different disciplines have different research priorities. This paper focus on spatial attributes of population mobility, the time attribute two attribute spread. Spatial attribute and human moving physical space corresponding to (eg, indoor and outdoor), the characteristic time attributes of the crowd moving time changes. According to the spatial and temporal information population activity get massive population movement trajectory, by learning the trajectory of specific people ,we can get its mobile activities based on the rule of information, mobile path prediction of this information of different kinds of people to explore the traffic planning, urban planning and control of influenza specific problem has a strong reference value. To build a conceptual model based on the individual, the study population distribution in local region and the flow law, and predicting user next stop moving, we design and implement the acquisition and analysis model of pedestrian trajectories, and the use of memory mobile Markoff chain (n-MMC) to predict the crowd moving track, this paper introduce the theory of these models foundation and verifying implementation, and application prospect of the prediction model are discussed.