城市交通网多车类备用能力模型与算法

考虑城市混合交通中公交车流的特殊性而建立了最优信号控制下的多车类备用能力模型来研究不同车类交通流之间的相互作用及对路网备用能力的影响．设计了带极值扰动的简化粒子群求解算法（dsP-SO），模型约束的处理采用边界附近不可行解部分保留的方式，给出的算例验证了该算法求解约束双层规划模型的有效性．研究结果表明，起--讫点（OD）总流量的增减不意味着该OD上所有车类流量都随之增减，设置适当的最小OD需求量乘子能确保各车类用户的利益不受损害．

Multiple vehicle classes reserve capacity model and algorithm for urban transportation network

In view of the particularity of public transit vehicles in mixed urban traffic flow, this paper proposed a multiple vehicle classes＇ reserve capacity model to research the interaction of different classes＇ traffic flow and the influence on network reserve capacity. This model was solved by a solution algorithm based on extremum disturbed and simplified Particle Swarm Optimization（dsPSO）, some infeasible solutions around constraints boundary were retained to guarantee the diversity of particle swarm in dsPSO iterative process. The results from numerical experiment prove the feasibility and validity of dsPSO in dealing with the constrained bi-level programming model, and it shows that not all classes of traffic flow increase or decrease with Origin-Destination（OD） flows, properly setting minimum OD demand multipliers can ensure that the profitability of all vehicle classes not be harmed,