Abstract: At present, the reliability analysis based on the stochastic traffic capacity is mainly achieved by searching the bottleneck of the road section, but for a long road section without a defined bottleneck, the stochastic analysis result of the capacity often has a large error.It is found that Weibull distribution as the distribution function of stochastic traffic capacity is more reliable, but its parameters may be affected by many factors.Based on the above problems, the mutation probability was introduced to explain the reliability evaluation method of expressway section composed of multiple detection positions. The proposed model could give the theoretical average capacity and mutation probability of different length sections in different time periods according to the requirements. At the same time,the quantitative relationship between parameters and influencing factors was given by studying the relevant data of the expressway in Zhejiang plain area.Finally,it was found that the risk of mutation of traffic flow in expressway section or road network could be well analyzed by examples.

Key words: traffic engineering, stochastic traffic capacity, Weibull distribution, reliability, mutation probability

摘要： 目前基于随机通行能力的可靠性分析主要是通过搜寻路段瓶颈来实现的，而对于某一未定义瓶颈的较长路段而言，通行能力随机性分析结果往往存在较大的误差。研究发现：以韦布尔分布作为随机通行能力的分布函数更具可靠性，但其参数可能受多种因素的影响。基于上述问题，引入突变概率来解释由多个检测位置组合而成的快速路段可靠性评估方法；该模型可按要求给出不同时段、不同长度路段的理论平均通行能力及突变概率；同时通过研究浙江平原地区快速路的相关数据，给出了参数与影响因素的定量关系。最后通过实例证明：该模型能较好地分析快速路路段或路网内交通流突变风险。

关键词: 交通工程, 随机通行能力, 韦布尔分布, 可靠性, 突变概率