轨道交通单线接运电动公交调度优化模型

Scheduling Optimization Model of Single Line Rail Transit Electric Feeder Bus

针对乘客由轨道交通换乘接运电动公交过程中等待时间长、候车成本高等问题,提出一种面向高峰时段乘客换乘情况的轨道交通接运电动公交的时刻表优化方法。通过分析换乘过程,以换乘乘客等候时间成本、换乘失败成本、电动公交使用成本和充电成本共4项成本之和最小为目标函数,以电动公交的发车顺序、换乘乘客的等待意愿、电动公交充放电特性对行驶里程产生的影响等作为约束条件,构建混合整数非线性规划模型。在接运公交的运输需求方面,考虑了除换乘乘客外本地乘客出行需求变化对接运电动公交时刻表的影响。最后提出一种混合人工蜂群算法求解模型,通过与遗传算法、粒子群算法的对比,进行了算法的敏感性分析。结果表明:目标函数总成本为1355.32元,相比原成本降低了23.56%,其中,换乘乘客等候时间成本为298.17元,换乘失败成本为84.03元,公交公司运营成本为867.40元,电动公交充电成本为105.71元,验证了构建的模型对时刻表优化问题的有效性。

Passengers may encounter some problems such as long waiting time and high waiting cost when they transfer from rail transit to electric feeder bus.A timetable optimization method for electric feeder bus was proposed,which was oriented to passenger transfer in rush hour.By analyzing the transfer process,the minimum sum of four kinds of costs,namely the waiting time cost of transfer passengers,transfer failure cost,using cost and charging cost of electric buses,was taken as the objective function,and the departure order of electric bus,the waiting willingness of transfer passengers,and the influence of electric bus charging and discharging characteristics on the driving distance were taken as the constraint conditions.A mixed integer nonlinear programming model was constructed.In terms of transportation demand for feeder bus.The impact of changes in the travel demand of local passengers other than transfer passengers on the electric feeder bus schedule was also considered.Finally,the model was solved by using the hybrid artificial bee colony algorithm,and the sensitivity analysis of the algorithm was carried out by comparing with genetic algorithm and particle swarm algorithm.The results show that the total cost of the objective function is reduced by 23.56%compared with the original cost.The total cost is￥1355.32,among which the waiting time cost is￥298.17,the transfer failure cost is￥84.03,the bus company operating cost is￥867.40,and electric bus charging cost is￥105.71.The validity of the constructed model for the timetable optimization problem is verified.