提高公交线路稳定性的公交专用道最优位置确定
Identify the Optimal Locations of Dedicated Bus Lanes to Improve the Stability of Bus Line
摘要：
公交车在线路上的扎堆和不均匀分布，会导致线路服务水平低下，影响线路的运行稳定性。为了解决上述问题，已经提出了作为控制手段的速度调节。但由于混合交通中车速调整的困难以及公交专用道的不连续分布，使得现有的公交车速度协调或调整方法在实际应用中遇到了严重的障碍。为了克服上述问题，我们首先考虑了沿巴士线部署后海湾线对速度调整策略性能的影响。以公交线路运营资金限制和对其他道路交通影响最大为主要约束条件，建立了以提高公交线路运营整体稳定性为目标的非线性混合整数随机数学模型。该模型可以有效地解决了一个新设计的分支定界算法。DBL中的最佳调节速度是通过一个向前看多个临界时间点的过程来确定的。数值实验验证了新方法的有效性和效率。获得了以下三个发现。为了确定最佳的调节速度，要向前看的临界时间点的数量不应该太大。增加DBL的数量提高了总线线路的稳定性。DBL沿总线分布得越均匀，总线的稳定性就越大。
英文摘要：
Bus bunching and unevenly dispersed buses along a bus line lead to a low service level and deteriorate the operational stability of the bus line. Speed adjustment as a control means has been proposed to solve the above problem. But due to the difficulty of adjusting speed in the mixed traffic and the discontinuously distributed dedicated bus lanes (DBLs) in a bus line, the existing methods which coordinate or adjust bus speeds in all road segments encounter serious obstacles in practice. To overcome the above problem, we first take into account the influence of the deployment of the DBLs along a bus line on the performance of the speed adjustment strategy. With the funding limit and the maximally allowable influence on other road traffic as main constraints, a nonlinear mixed integer stochastic mathematic model is formulated aiming at improving the whole stability of the bus line operation. The model can be effectively solved by a newly designed branch and bound algorithm. The optimal regulating speed in a DBL is determined with a process of looking multiple critical time point ahead. Numerical experiment verified the effectiveness and efficiency of our new method. Three findings were obtained as follows. To determine the optimal regulating speed, the number of the critical time points to be looked ahead should not be too big. Increasing the number of DBLs improves the stability of a bus line. The more evenly DBLs are distributed along the bus line, the greater the stability of the bus line will be.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
