摘要
在考虑客流到达和服务随机性的基础上,首先验证了理论上可以逼近任意分布的PH分布能很好地拟合客流到达和服务规律,然后建立了描述地铁候车排队现象的PH/PH^[0,C1]/1/C批量服务排队模型,并采用AnyLogic软件仿真验证了模型的通用性和准确性。结果表明,针对9组不同到达率和变异系数的客流数据,仿真模拟和排队模型计算的各项指标的最大绝对误差在0.01～6.06之间,最大相对误差在0.91%～7.69%之间,并且模型结果均在仿真结果的95%置信区间内,故该模型可较好地描述地铁候车排队现象。
With a full consideration of the potential passenger arrival and the randomness of service,the PH（hydrogenion concentration）distributions that can fit well with the law of potential passengers and services are demonstrated in theory.Then,the related model that describes metro waiting queues of PH/PH^[0,C1]/1/C is established.In which,AnyLogic software simulation is used to analyze and verify the versatility and accuracy of the model.The results indicate that the maximum absolute error of the calculated simulation and queuing model indicators is between 0.01～6.06,based on 9 different set arrival rates and variation coefficient of the passenger flow data,and the relative maximum error is between 0.91% ～7.69%.Thus,the model results are in 95% confidence interval in the simulation results,thus verify the reliability of this model.
出处
《城市轨道交通研究》
北大核心
2016年第7期9-14,共6页
Urban Mass Transit
基金
国家自然科学青年基金项目(51108391)
西南交通大学优秀博士学位论文培育项目
