欢迎访问《汽车安全与节能学报》,

汽车安全与节能学报 ›› 2022, Vol. 13 ›› Issue (2): 341-349.DOI: 10.3969/j.issn.1674-8484.2022.02.015

• 汽车节能与环保 • 上一篇    下一篇

基于聚类与Markov链法的西安市某线路城市客车工况构建

李耀华(), 邵攀登, 翟登旺, 任田园, 宋伟萍, 刘洋, 赵承辉   

  1. 长安大学 汽车学院,西安 710064,中国
  • 收稿日期:2021-01-18 修回日期:2022-04-01 出版日期:2022-06-30 发布日期:2022-07-01
  • 作者简介:李耀华(1980—),男(汉), 陕西, 副教授。E-mail: nuaaliyaohua@126.com
  • 基金资助:
    国家自然科学基金资助项目(51207012);陕西省自然科学基金资助项目(2021JM-163)

Construction of driving cycle of city bus line in Xi’an based on clustering and Markov chain method

LI Yaohua(), SHAO Pandeng, ZHAI Dengwang, REN Tianyuan, SONG Weiping, LIU Yang, ZHAO Chenghui   

  1. School of Automobile, Chang’an University, Xi’an 710064
  • Received:2021-01-18 Revised:2022-04-01 Online:2022-06-30 Published:2022-07-01

摘要:

为了对特定区域构建符合当地车辆行驶特征的行驶工况,基于聚类与Markov链法构建了西安市某线路城市客车的行驶工况,确定了聚类个数及特征参数组合,提出了构建工况长度的确定方法,从能耗角度定义汽车行驶时的单位里程比能耗作为工况选取标准,从50条候选工况中筛选出该线路的代表工况。结果表明:与聚类法工况和V-A矩阵法工况相比,基于聚类与Markov链法构建的行驶工况与样本数据偏差最小,平均偏差率为1.17%,百千米能耗相差最小,偏差率为0.069%,显示基于聚类与Markov链法构建的行驶工况精度更高,更能反映车辆的实际行驶状况工况。

关键词: 聚类法, Markov链法, 行驶工况, 能耗

Abstract:

A driving cycles of Xi’an city bus were constructed by clustering and Markov chain method to construct driving conditions that reflect the local vehicle driving characteristics for a specific area. The number of clusters and characteristic parameters in clustering were determined, and the method to determine the length of driving cycle by Markov chain was proposed. The specific energy consumption per mileage was defined as the criterion to select typical driving cycle among 50 candidate driving cycles. The results show that, compared with the clustering method and the V-A matrix method, the driving conditions constructed based on the clustering and Markov chain method have the smallest deviation from the sample data, the average deviation rate is 1.17%, and the difference in energy consumption per 100 kilometers is the smallest, the deviation rate is 0.069%. The driving cycle constructed by clustering and Markov chain method shows higher accuracy and can reflect the actual driving conditions better.

Key words: clustering method, Markov chain method, driving cycle, energy consumption

中图分类号: