大数据环境下基于人类动力学的民航交通分布模型研究

项目名称： 大数据环境下基于人类动力学的民航交通分布模型研究

项目编号： No.U1333113

项目类型： 联合基金项目

立项/批准年度： 2014

项目学科： 无线电电子学、电信技术

项目作者： 彭舰

作者单位： 四川大学

项目金额： 36万元

中文摘要： 民航交通分布与旅客出行密切相关，现有交通分布模型未考虑旅客出行的规律性和可预测性。人类动力学研究表明，人类空间移动具有高度规律性和可预测性，这种可预测性高达93%，从人类动力学角度研究民航交通分布体现问题本质。本课题将人类动力学应用于交通分布建模，主要内容包括：（1）采用大数据预处理方法获取精简规范数据；（2）采用参数估计与假设检验方法实证人类空间移动无标度特性在民航出行中的普适性；（3）采用相关性分析等方法研究人造基础设施网络对人类空间移动的影响；（4）利用机器学习、排队论并结合已有建模方法建立结合民航网络的人类空间移动模型；（5）利用统计物理、复杂网络等理论进行群集动力学建模，建立基于人类动力学的民航交通分布模型。本课题实证人类空间行为动力学两个基础性问题，首次将人类动力学应用在民航交通分布建模，具有重要的理论意义和应用价值。

中文关键词： 人类动力学；航空旅客；移动模型；大数据；数据挖掘

英文摘要： Civil aviation traffic distribution is closely related to individual travel, but the existing traffic distribution models did not take the regularity and predictability of individual mobility into consideration. Research on human dynamics shows that human mobility is characterized by a deep-rooted regularity and predictability, and a 93% potential predictability has been found in human mobility. In this project, we apply human dynamics to modeling the civil aviation traffic distribution, and our research includes : (1) to get the compact and standardized data by using the method of the big data preprocessing; (2) to do an empirical demonstration research on the universality of scale-free property of individual civil aviation travel through the methods of parameter estimation and hypothesis testing; (3) to study how the artificial infrastructure network influence on the human mobility by correlation analysis and other methods; (4) to establish the human mobility model based on airline network by applying machine learning, queuing theory combined with existing modeling methods; (5) to establish the civil aviation traffic distribution model based on human dynamics through employing the multi-agent technology to simulate the group mobility and applying theories including statistical physics, complex networks to dyna

英文关键词： human dynamic；air passenger；mobility model；big data；data mining