随机环境下公交车头时距自均衡控制及公交运力鲁棒优化方法

Public transit system, influenced by the random disturbance in the process of running, cannot accurately follow the fixed operation scheduling plan, leading to the fleet size required larger than it in ideal environment, increasing the cost of buying buses and operations, inducing more congestion on the urban road. The project, to relieve bus bunching and its collateral issue redundancy of fleet size, based on the support of short term transit passenger flow forecast data, using the means of coordination of bus holding and skip stop, with self-adaptive equalizing headways control strategy, proposes robust optimization method of public transport in random environment in format of analytical model that bus system can automatically recover to a stable state after disturbance. Furthermore, considering the self-adaptive equalizing headways control method, the cost and risk objective functions are established by analyzing the relationship between real time bus control and bus fleet size optimization, obtaining the optimization bus fleet size under certain service level. The study can enrich the theory of urban public transport intelligent operation system, and theoretical system of optimization for bus fleet size, having high theoretical value and practical significance.

城市道路交通包含大量随机扰动，受此影响公交系统难以维持运营稳定性，致使公交运力规模需求增加，一方面造成公交车购置和运营成本增加，另外也加大了对城市道路交通的压力。为此，针对公交串车及公交车辆冗余问题，本项目以短时公交客流预测数据为基础，以公交驻站和公交跳站为协调控制手段，以公交车头时距自均衡为控制策略，以模型预测控制为算法基础，以解析模型为最终形式，构建随机环境下公交车头时距自均衡优化控制模型，通过对控制点处公交车实时控制使公交车头时距在受到扰动后能够自动恢复至均衡状态；在此基础上，以车头时距自均衡为底层控制方法，以公交系统成本和运营风险为目标函数，统筹分析动态公交实时控制与静态公交运力规模之间的内在联系，提出基于鲁棒优化思想的公交运力规模优化模型，实现在一定服务水平下公交线路上车辆数最优。项目研究成果可丰富城市公共交通控制和公交运力规模优化理论体系，具有较高的理论价值和实际意义。

本项目针对城市公交系统随机扰动诱发的公交无序运营问题，以常规公交系统频发的公交串车为研究对象，针对公交客流到达公交站点的随机波动性，以二维时间序列上公交站点的客流数据统计与分析为基础，设计了基于二维时间序列的公交站点客流预测模型，进而以公交驻站和公交跳站协调控制为手段，以自均衡车头时距为控制策略，设计了随机环境下公交系统鲁棒性优化方法，提升公交系统抵御随机干扰的能力，在此基础上对公交规模进行优化，在满足一定服务水平的基础上使得公交车队规模最小。最后，对项目中所提出的客流预测方法、公交车头时距协调控制方法和公交车队规模优化方法分别进行了实验验证，根据实验结果显示，本项目所提出的方法均具有较好的性能，能够较好的满足工程实践需要。