中文摘要：

本文在传统有限马尔科夫链的随机过程中，加入时间因素和空间属性，构造时空马尔科夫链，并以北京市2011年人口密度、林木绿化率、学校密度、每千人拥有医院床位张数为基期数据，预测至2020年北京城乡人口流动状态和趋势。研究发现：在2018年以前，北京市城乡人口流动状态是不稳定的，城乡转移概率曲线呈现先凸后凹的雁阵特征，空间分异趋势明显；至2018年，北京市城乡人口迁移概率趋于一种相对稳定的状态，但城乡转移概率的差距并未缩小，城乡二元格局在短期内不会改变。因此，为缓解特大城市资源环境承载压力，京津冀三地政府应联合制定省际经济利益协调机制，实行区域经济政府政绩一体化考核，加快新城建设，发展功能互补的中小城镇，合理引导产业转移，以业带城，形成城市群，以小城镇赢大战略，均衡配置公共资源，统筹城乡空间布置，放开卫星城户籍限制，实施城乡医疗保险一体化改革。

英文摘要：

This paper builds space-time Markov chain based on the traditional finite Markov chain and together with time factor and space attribute. This paper predicts he state and trend of the population mobility between rural and urban areas in Beijing till 2020 according to the population density, forestation rate, school density, the number of bed in hospital per one thousand people in 2011. The research discovers that the state of mobility between urban and rural areas in Beijing will be unstable and the curve of transfer rate between urban and rural areas will present the formation of flying swan goose before 2018. The transfer rate between urban and rural areas in Beijing will have been stable till 2018, but dual structure of urban and rural areas in Beijing will have not been changed. Therefore, the local government ought to set up the coordination mechanism for interprovincial economic interest of Beijing, Tianjing and Hebei, innovate performance evaluation for integration, build up new town as soon as possible, promote industry to transfer rationally, advance the development of city cluster, balance public goods, make overall plans for urban and rural spatial stricture, break the constraint about the household registration and combine the system of rural medical insurance with the system of urban medical insurance.