摘要
利用实测的公交车运行数据,建立符合公交车运营特点的行驶工况。首先将连续行驶数据进行短行程划分并计算各短行程的特征值;之后采用主成分分析将12个特征值降为4个主成分,利用相关系数法建立了武汉市公交车的综合行驶工况;同时采用聚类分析对短行程分类,构建了公交车在拥堵道路、较畅通道路、畅通道路3类交通条件下的行驶工况;各工况同实测数据的相关系数均超过了0.98。研究结果表明,该地区公交车平均运行速度为19.46km/h,各行驶模式下的时间比例分别为:加速26.39%、减速23.61%、匀速33.33%、怠速16.67%。此外将所建立综合工况与燃油消耗量测试工况C-WTVC比较,发现二者在平均速度和怠速时间比例方面存在较大差别。因此有必要针对公交车专门开发测试工况,从而为交通和环保部门的公交运营管理提供指导。
The measured driving data of public buses were utilized to establish driving cycles that conform to bus operating characteristics. First, the continuous driving data were divided into multiple micro trips. Then, the characteristic values of each trip were calculated. Furthermore, the characteristic values were compressed from twelve values to four principal components by using the method of principal component analysis. The synthesized driving cycle for public buses in Wuhan was constructed by utilizing the correlation coefficient method. Meanwhile, the driving cycles of public buses under three different traffic conditions (congested road, uncongested road, and light traffic road) were constructed by using the clustering analysis method. The correlation coefficients of each driving cycle with the measured data all exceed 0. 98. The results reveal that the average speed of public buses in Wuhan is 19.46km/h and the time proportions in four driving modes are:acceleration 26.39%, deceleration 23.61%, cruise 33.33%, and idle 16.67%. Moreover, the results of comparison between the established synthesize driving cycle and the fuel consumption test cycle of C-WTVC reflect that their average speed and idle time percentage have bigger differences. Therefore, it is necessary to construct specific test cycles for public buses. The test can provide guidance for the bus operation management of traffic and environmental protection agencies.
出处
《交通信息与安全》
2014年第6期139-145,共7页
Journal of Transport Information and Safety
基金
国家自然科学基金项目(批准号:51478045)
中国博士后科学基金项目(批准号:2013M532006)
陕西省科技统筹创新工程项目(批准号:2012KTZB03-01-01)资助
关键词
交通工程
行驶工况
主成分分析
聚类分析
公交车
交通条件
traffic engineering
driving cycle
principal component analysis
clustering analysis
bus
traffic condition