At present, the existing train operation plan optimization research is based on the train stops in every station.In addition, in these researches, the stop time of the train is fixed, and the number of the passengers during the optimizing period is also fixed, and at the same time, only one-way operation optimization is considered.Firstly, the stopping time of the train should be set as a dynamic value, then the arrival rate distribution of the passengers should be considered, the shortest traveling time of the passengers and the lowest enterprise production cost should be regarded as the target function.The passenger train waiting penalty coefficient felt by the passengers should be introduced to structure the multi-objective programming model for station-crossing train non-equilibrium operation plan based on dynamic station stop time, and design the solve process based on genetic annealing algorithm.Finally, the rail transit line of a city is analyzed as the example, the result shows that, compared with the scheme of train stopping in every station with fixed stopping time, this method can save about total 6.5％travel time for the passengers and 3.7％operating costs for the enterprise, so that the optimizing effect is quite obvious.%针对当前已有列车开行方案优化研究大多以站站停为前提,且固定列车停站时间、将优化时段内旅客人数设为定值,同时只考虑列车单向运营优化等不足.首先,将列车停站时间设为动态值,接着考虑乘客到达率分布,以乘客总出行时间最小及企业生产成本最低为目标函数,引入考虑乘客心理感知的乘客候车惩罚系数,构造了基于动态停站时间的跨站列车非均衡开行方案多目标规划模型,并设计了基于遗传退火算法的求解流程.最后,以某市城市轨道交通线路为算例进行分析,结果表明,与固定停站时间的站站停方案对比,该方法可节约乘客总旅行时间6.5％、企业运营成本3.7％,优化效果较为明显.
展开▼
