Characteristic Analysis And Modeling Of Human Mobility Behavior Between Cities

Understanding how people move between different cities is not only a basic problem in many subjects,such as traffic science,economic geography,demography,but also has important applicative value in many other fields.For example,the short-term moving behavior analysis on people between cities(i.e.city travel behavior)helps us understand how infectious diseases spread at large spatial scales.In addition,long-term migration behavior analysis of intercity populations has important significance in the study of city development and the formulation of regional development policy.With the rapid development of modern communication technology and Internet,more and more data can be used to record the mobile footprints of large populations,which provides a new opportunity for the study of crowd mobility.However,traditional data such as household registration,which can reflect the population migration between cities,is sometimes difficult to obtain for the general researchers.Thus,the limits restrict our extensive research on the law of population migration between cities.Recently,researchers have studied the migration behavior of scholars in cities by using the authors' work addresses recorded in the scientific literature database.Unfortunately,due to restriction on data scale,limitations of research objects and other reasons,we still lack enough understanding of the migration law of the general population in the city.In this paper,the migration behavior of large-scale and multi occupational people in different cities of China is studied by using anonymous job seekers' information crawled from a recruitment website in China.We can find that long-term intercity traffic structure networks has a highly asymmetric characteristics differed from the short-term mobile network,which reflects the difference of populations and competition effect of city attraction.This flow structure can be approximately described by the gravity model of spatial economics.And the value of scaling exponent of distance function in the gravity model is less than the value of short-term travel network.It means that long-term human migration behavior is less sensitive to the distance than short-term travel behavior.In addition,we compare the scaling coefficients of each variable in the gravity model and find out that the economic level has a greater influence on the migration population.Combined with the data of urban economic development level,this paper uses an improved population-weighted opportunity model to reproduce and explain the above modes of intercity migration network.In addition,this paper uses a gravity model to predict the human migration behavior with the data of economic quantity and population.And the difference of the index of different parameters is analyzed and compared.Finally,theories and models are used to systematically compare and explain the varieties of the related parameters between urban migration behavior and travel behavior.