关于公交车调度的优化问题

Optimization of Dispatching Buses

本文主要是研究公交车调度的最优策略问题。我们建立了一个以公交车的利益为目标函数的优化模型 ,同时保证等车时间超过 10分钟 (或者超过 5分钟 )的乘客人数在总的等车乘客数所占的比重小于一个事先给定的较小值α。首先 ,利用最小二乘法拟合出各站上 (下 )车人数的非参数分布函数 ,求解时先用一种简单方法估算出最小配车数 43辆。然后依此为参照值 ,利用Maple优化工具得到一个整体最优解 :最小配车数为 48辆 ,并给出了在公交车载客量不同条件下的最优车辆调度方案 ,使得公司的收益得到最大 ,并且乘客等车的时间不宜过长 ,最后对整个模型进行了推广和评价 ,指出了有效改进方向。

It is to find out the best way to dispatch buses. We set a optimized model whose target function is the profit of bus company.At the same time, it guarantee the proportion that the passengers waiting for their buses more than 10 min (or 5 min)in the total is less than α given before.First, every station's nonparameter distribution function about the number of passengers is fitted by method of least squares. We use a simple method to estimate that at least 43 buses are needed, and then, we use Maple to get the optimal solution refer to it. It shows the best plans for dispatching buses in different conditions of the number of passengers. It can help bus companny to get the top profit, meanwhile the passengers may not wait for their bus for a long time. In the end, we evaluate and popularize the model, and point out the effective way to improve it.