针对地铁列车跨线运行模式乘务排班和轮班计划协同优化的问题，研究循环轮班模式一体化优化对乘务计划编制效率和乘务员运用效率的影响。基于循环轮班班制构造时空网络搜索乘务员在轮班周期内的乘务区段和班次接续顺序，以乘务员值乘班次数量和值乘空闲时间最小化为优化目标构建模型；构造乘务员班次接续约束和班次可行性约束限制乘务员的值乘路径，保证在满足乘务区段接续规则的前提下确定轮班周期内乘务员的值乘任务；基于班次出退勤轮乘站、班次类型、轮班周期、轮班班制设计轮班路径搜索算法和改进列生成算法获取轮班周期内的值乘安排，提出混合班制的乘务轮班模式探讨四班三运转和六班五运转班制混合对乘务计划的影响。结果表明：相较传统四班三运转和六班五运转轮班模式，混合班制可使一体化轮班模式班次的平均工作效率分别提高1.5和2.3个百分点，使便乘区段数分别降低12.18%和24.45%；相较分阶段优化，在不影响乘务员值乘班次数的基础上，一体化优化提高了班次平均工作效率和乘务员运用率，降低了轮班周期内的总班次数和乘务区段冗余覆盖程度。混合班制模式下的一体化优化方法可适应灵活的轮班周期和各线路乘务区段时空分布的差异性，有利于保障乘务员的值乘均衡性和运用效率。

This study addresses the collaborative optimization of crew scheduling and rostering for subway trains operating in a cross-line pattern. It investigates the impact of an integrated optimization approach, based on a cyclic roster system, on both the efficiency of crew plan formulation and the utilization rate of crew members. A spatio-temporal network is constructed based on the cyclic roster system to search for feasible crew duty segments and the sequence of shift connections within a roster cycle. A mathematical model is formulated with the dual objectives of minimizing the total number of crew shifts and minimizing idle time during duties. Constraints are established, including crew shift connection rules and shift feasibility requirements, to define permissible roster paths for crew members. These constraints ensure that the assigned duties within the roster cycle comply with the rules for connecting duty segments.A roster path search algorithm and an improved column generation algorithm are developed, considering factors such as sign-on/sign-off depots for shifts, shift types, roster cycle length, and the shift system design. These algorithms are employed to obtain optimal duty assignments within the roster cycle. Furthermore, a hybrid roster system is proposed, exploring the effects of mixing “four crews for three operational shifts” and “six crews for five operational shifts” systems on the crew schedule. The results show that, compared to the traditional separate “four crews for three shifts” and “six crews for five shifts” systems, the proposed hybrid system within the integrated optimization framework increases the average shift efficiency by 1.5 and 2.3 percentage points, respectively. It also reduces the number of deadhead segments by 12.18% and 24.45%, respectively. Compared to a sequential (two-stage) optimization approach, the integrated method improves average shift efficiency and crew utilization rate without increasing the number of shifts worked per crew member. Additionally, it reduces the total number of shifts required within the roster cycle and decreases the redundancy in covering duty segments. The integrated optimization approach with the hybrid roster system can adapt to flexibile roster cycles and the spatio-temporal distribution differences of duty segments across various lines. This adaptability is beneficial for ensuring balanced duty assignments for crew members and enhancing their overall utilization efficiency.