Scholarlink

<p>Slopes often emerge as traffic bottlenecks, yet not all slopes lead to congestion. The relationship between slope capacity and factors like grade and length is complex and non-linear. Accurately estimating road slope capacity and mitigating traffic congestion remain challenges in traffic management. Drivers instinctively adjust vehicle acceleration or braking to counteract gravity, influencing vehicle speed and road capacity. However, traditional models often overlook these compensatory adjustments, leading to inaccurate predictions. This study introduces a novel approach using vehicle trajectory data and the Expectation-Maximization (EM) algorithm to estimate driver compensations on slopes. The algorithm separates observed acceleration into baseline (flat road) and compensatory components. Data from field experiments with cars, trucks, and buses reveal that compensatory acceleration decreases with speed and remains predictable across different slopes. These findings enhance our understanding of slope impacts on traffic flow and provide valuable insights for traffic management and infrastructure design.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Stochastic driver compensation for road gradients across vehicle classes: a machine learning approach

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年5月28日

查看全文
  • Kai Yuan,
  • Hongye Xing,
  • Weihua Zhang,
  • Rui Jiang,
  • Shixiong Jiang

摘要

Slopes often emerge as traffic bottlenecks, yet not all slopes lead to congestion. The relationship between slope capacity and factors like grade and length is complex and non-linear. Accurately estimating road slope capacity and mitigating traffic congestion remain challenges in traffic management. Drivers instinctively adjust vehicle acceleration or braking to counteract gravity, influencing vehicle speed and road capacity. However, traditional models often overlook these compensatory adjustments, leading to inaccurate predictions. This study introduces a novel approach using vehicle trajectory data and the Expectation-Maximization (EM) algorithm to estimate driver compensations on slopes. The algorithm separates observed acceleration into baseline (flat road) and compensatory components. Data from field experiments with cars, trucks, and buses reveal that compensatory acceleration decreases with speed and remains predictable across different slopes. These findings enhance our understanding of slope impacts on traffic flow and provide valuable insights for traffic management and infrastructure design.