基于贝叶斯推理的HCM延误模型修正

摘要

针对以1个周期时长为分析单位、使用HCM2000延误模型推导信号控制交叉口延误的问题,提出推导模型中参数修正的方法,用t检验验证参数提取的精度.对延误提取模型中的饱和度、启动损失时间及交叉口几何修正系数等参数进行分析,采用贝叶斯定理和马尔科夫链蒙特卡罗模拟方法对参数进行修正.结果证明该方法可以提高按照周期提取延误参数的精度.%This paper validates the HCM2000 delay model at the normal traffic condition using 1 cycle as the duration analysis period. The precision of the HCM 2000 delay model is tested using the t-test method. The saturation, geometry parameter and the start-up loss time parameter in the model are analyzed. Bayesian inference and Markov Chain Monte Carlo(MCMC) simulation are used to revise the models parameters. The method can improve the model's precision when using 1 cycle as the duration of analysis period.