考虑超载控制与内生拥挤的长短线公交动态运营策略

Dynamic Mixed Fleet Operation Strategy for Feeder Service Considering Overload Control and Endogenous In-vehicle Congestion

针对长线公交在部分时段和局部线路上的拥挤,本文提出在部分时段增发短线公交的运行策略。考虑时变的出行需求,本文构建了带有均衡约束和机会约束的动态规划模型,通过近似和逆推解法对动态规划模型进行快速求解,以确定短线公交的动态发车方案。一方面,本文构建机会约束控制车内拥挤,从而帮助提高公交的服务水平,这响应了出行者对公交服务水平的要求越来越高的现实;另一方面,分析了内生拥挤、客流在长短线之间的转移,使得本文的研究更贴合实际。考虑到动态策略的操作难度,本文建立混合整数规划模型对动态策略的启发式特例进行求解。算例分析结果表明,本文提出的动态策略和启发式策略均能通过增加适量的运营成本,有效地缓解长线公交在高峰时期的超载,同时显著减少乘客候车时间成本和车内拥挤;启发式策略能在乘客体验和运营成本之间进行合理权衡,并且在操作简易性方面更具优势。

This paper proposes a mixed fleet strategy for the high-demand feeding section of a long-distance bus line to alleviate its in-vehicle congestion and reduce passengers’waiting time during peak hour.The strategy dynamically dispatches additional feeder buses with different capacities.To determine the optimal dispatching scheme,a dynamic program with equilibrium and chance constraints is established,wherein the equilibrium constraints are used to capture the demand shift between the long-distance bus and its following feeder bus while the chance constraints are used to control the overload probability.The model can be efficiently solved through approximation and backward induction.A heuristic special case of the dynamic strategy is also developed through mix-integer programming.Numerical results show that both the dynamic strategy and the heuristic can reduce the in-vehicle congestion,overload risk and passengers’waiting time.The heuristic strategy can strike a desirable balance between passenger experience and operational cost.