基于轮轨定位数据的有轨电车区间驾驶特征分析
Analysis of Tram Section Driving Characteristics with Wheel Odometer Data
童文聪 1滕靖 1李君羡 1姚幸 2张中杰2
作者信息
- 1. 同济大学 上海市轨道交通结构耐久与系统安全重点实验室,上海 201804
- 2. 上海市有轨电车工程技术中心,上海 200125
- 折叠
摘要
为分析人工驾驶条件下有轨电车区间速度及可靠性特征,基于轮轨定位数据,计算有轨电车在加速段、巡航段、制动段和交叉口的运行特征指标,分析人工驾驶决策对各指标的影响机制;并建立区间运行速度的多因素回归分析模型及概率分布模型.结果表明:由于人工驾驶的模糊控制特点,司机无法实现充分加减速;终点速度和制动系数对区间运行速度贡献度总占比达57%,是驾驶行为优化的重点;区间运行速度呈高斯混合分布(Gaussian Mixture Model,GMM),对常见绿波带宽有较高的偏出率,是造成线路时间可靠性低的重要原因.
Abstract
To analyze the speed and reliability characteristics of trams under manual driving conditions in sections,wheel odometer data was used to calculate the characteristic indexes in acceleration,cruise,braking and intersection section,and influence mechanism of manual driving decision on each index was proposed.Multi-factor regression model and probability distribution model of section running speed are established.The results indicate that driver can not achieve full acceleration and deceleration due to the fuzzy control of manual driving;the contribution of terminal speed and braking coefficient to section running speed accounts for 57%,which is important for driving behavior optimization;the section running speed conforms to Gaussian mixture distribution(GMM)with high deviation rate to the common green wave bandwidth,resulting in low time reliability of tram lines.
关键词
有轨电车/驾驶行为/速度特征/多元线性回归/高斯混合分布Key words
tram/driving behavior/velocity characteristics/multiple linear regression/Gaussian Mixture Model引用本文复制引用
基金项目
国家重点研发计划(2021YFB1600100)
出版年
2024
国家科技期刊平台
NSTL
万方数据