对城市轨道交通动态可达性及末班车时刻表的研究，是优化轨道交通末班车换乘衔接，保证运输效益最大化的基础． 首先，从点、线、面三个层面，在综合考虑路网静态拓扑结构及动态 OD(Origin-Destination)客流的基础上，提出城市轨道交通动态可达性指标及其量化方法;其次，面向广州市轨道交通网络，基于 OD 客流需求进行网络客流加载并评价轨道交通网络可达性;最后，以动态可达性最大为目标，考虑各线路发车时间等约束，建立末班车时刻表优化模型，并以广州地铁为例，基于遗传算法对其末班车时刻表进行调整优化． 结果表明:调整后广州地铁路网末班车静态可达性提高 20. 29%，动态可达性提高 10. 97 个百分点，优化后的末班车时刻表能有效提高运输能力，该模型可为网络运营组织提供决策依据．

The research of urban rail transit dynamic accessibility and last train timetable is the basis for optimizing the transfer between last trains in different lines and maximizing transportation benefits． First，urban rail transit dy-namic accessibility is defined from three dimensions such as stations，transit lines，and the whole network，based on the static network topology and dynamic passenger flow． Second，dynamic accessibility indicators of Guangzhou rail transit network is evaluated through assigning the dynamic passenger flow data to the whole network． Finally，aiming at maximizing dynamic accessibility，a last train optimizing model is proposed，taking into consideration the departure time constraint of the last trains for all lines，and a genetic algorithm is applied to solve the optimization problem． Taking Guangzhou subway network for the case study，the analysis shows that static accessibility and dy-namic accessibility of Guangzhou rail transit network increased by 20. 29% and 10. 97percentage points respectively after the adjustment of the last train timetable，which indicates that the optimized last train timetable can effectively improve transport capacity．