传统居民出行调查数据缺陷长期影响交通规划实践效果、制约交通需求模型应用演进。4G/5G通信技术革新推进个体出行定位从“稀疏模糊”基站位置序列向“密集模糊”甚至“高频高精”坐标轨迹升级，为精细化个体出行信息采集带来空前契机，备受行业关注。因此，本课题致力于研究基于4G/5G连续密集定位轨迹的个体出行链信息精细化提取理论与方法。依托多场景实测与仿真试验，量化研究4G/5G通信网络追踪个体出行轨迹机理与效能，揭示4G/5G定位精度、信度与稳定性；探索4G/5G定位轨迹对个体出行链全局与局部特征的时空表达，应用时空聚类、深度学习、证析理论等建立个体出行链参数识别模型，重点突破出行端点、换乘点、交通方式和出行目的识别4大技术难点；研究出行链成链规则与成链模式，提出信息成链视角下出行链参数宏中微多维误差自检与协同优化方法。本研究可为交通管理部门提供个体出行调查新方法，为交通需求模型演进提供数据支撑。

The bottleneck of data quality caused by traditional residents' trip survey method has long affected the effectiveness of traffic planning and restricted the evolution of traffic demand analysis model. The innovative 4G/5G communication technology promotes the upgrading of individual travel trajectory detection from sparse-fuzzy base station sequence to dense-fuzzy or even high frequency and high precision coordinate trajectory, which brings opportunity for refined individual trip information collection and attracts industry attention. Therefore, this project focuses on the theory and method of refined extraction of individual trip information based on 4G/5G continuous dense positioning data. Relying on multi-scene measurement and simulation experiments, the mechanism and efficiency of 4G/5G positioning are studied quantitatively, and the accuracy, reliability and stability of 4G/5G positioning are revealed; By using advanced algorithms such as time-space clustering, in-depth learning and evidence analyzing, individual trip chain information detection models are proposed, breaking through the four major technical difficulties of trip end, mode transfer point, trip mode and trip purpose detection. The trip chain parameter link rules and modes are studied, a method of self-checking and collaborative optimization for macro, medium and micro trip information detection errors from the perspective of information chaining is proposed. This project can provide a new travel survey method for traffic management department and data support for traffic demand model innovation.