贝叶斯框架下快速公交的站间运行状态推断

发布时间：2019-02-15 19:06

【摘要】：站间行程车速是度量快速公交运营服务和乘客体验的重要指标.由于红绿灯、行人过街等影响,快速公交车速分布往往呈现多峰现象,很难利用单一分布描述.本文建立了贝叶斯框架下基于多相站间车速的高斯混合模型,利用可逆回跳马尔科夫链蒙特卡洛方法推断快速公交站间运行状态,包括状态个数和各状态权重、速度均值和方差.以广州市快速公交为例,展示了该模型可很好地拟合多相车速数据,合理划分快速公交运行状态.发现广州市快速公交在早晚高峰容易衍生出新的运行状态,对于车速低于25 km/h的走廊,其运行状态个数较多,且更容易随时段发生变化,符合实际调查结果.可见本文提出的模型能有效地分析快速公交的速度分布特征和演变规律,验证了该模型的有效性.
[Abstract]:Interstation travel speed is an important index to measure bus rapid transit service and passenger experience. Due to the influence of traffic lights and pedestrians crossing the street, the distribution of BRT speed often presents multi-peak phenomenon, which is difficult to describe by single distribution. In this paper, a hybrid model of Gao Si based on the speed between multi-phase stations is established in Bayesian framework. The reversible bounce-back Markov chain Monte Carlo method is used to infer the operation state of bus rapid transit stations, including the number of states and the weight of each state. Velocity mean and variance. Taking bus rapid transit in Guangzhou as an example, it is shown that the model can fit the multi-phase speed data well and reasonably divide the running state of bus rapid transit. It is found that the bus rapid transit in Guangzhou is easy to derive a new running state in the morning and evening rush hour. For the corridor whose speed is less than 25 km/h, the number of running states is more, and it is easier to change at any time, which is in line with the actual investigation results. It can be seen that the model presented in this paper can effectively analyze the speed distribution characteristics and evolution law of BRT, and verify the validity of the model.
【作者单位】：中山大学公共实验教学中心;广东省智能交通系统重点实验室;中山大学工学院智能交通研究中心;
【基金】：国家自然科学基金项目(41271181) 广东省环境监测大数据公共服务平台及其创新应用(2015B010110005)~~
【分类号】：U491.17

【相似文献】