基于回归分析的交织区通行能力模型

A Regression-Based Capacity Model for Weaving Areas

针对以往交织区通行能力回归分析中缺乏较为通用的回归模型、导致模型设计盲目性较大的问题,以HCM2000中交织区通行能力建议值为基础,建立模型的基本形式。研究表明,双曲线模型可较好地描述交织区通行能力与交织区长度的关系,模型参数可通过对交织区类型、交织流量比、交织区车道数和自由流车速等因素进行多元线性回归拟合得到。经检验,模型拟合度为0.918,平均误差为4.4%,说明模型可用且准确性较高。将模型应用于天津市快速路交织区通行能力计算中,验证了模型的实用性和可移植性。

Due to a lack of regression models generally used for weaving area capacity analyses, which leads to a blindness in model design, this paper develops a base model, based on the suggested values of weaving area capacity by HCM2000. Results show that the hyperbolic model yields a good illustration of relationship between weaving capacity and weaving length, and that the model parameters can be calibrated through a multi-linear regression by considering factors like weaving types, weaving flow ratios, number of weaving lanes, and free speeds. A statistic test shows the model is promising, with a good-fitness of 0.918, and an average error of 4.4%. A demonstration in capacity analysis of an expressway in Tianjin shows the model is applicable elsewhere as welt.