基于用户公共道路大数据的工况分类方法研究  

Research on Working Condition Classification Method Based on User Public Road Big Data

作　　者：刘泽奇 朱雄 王庆华 彭良峰 黄刚 LIU Ze-qi;ZHU Xiong;WANG Qing-hua;PENG Liang-feng;HUANG Gang(National Automobile Quality Inspection and Test Center(Xiangyang),Xiangyang 441004,China)

机构地区：[1]国家汽车质量检验检测中心(襄阳),襄阳441004

出　　处：《汽车科技》2024年第6期53-58,共6页Auto Sci-Tech

摘　　要：随着新能源汽车产业的快速发展,为提高车辆可靠性,降低车辆能耗和排放,准确评估和分析在实际应用中车辆的行驶状态至关重要。由于车辆工况测试模型的构建在很大程度上依赖于实际驾驶状态。提出一种基于用户公共道路大数据的车辆工况分类方法,针对不同地区、不同用户、不同工况构建符合实际驾驶特点的载荷谱采集方案。该方法对大量真实驾驶数据进行整理清洗、降维分析、聚类分析,最终为可靠性试验工况构建提供科学依据。结果表明:建立用户驾驶模型和车辆可靠性试验工况模型的关联,有助于验证车辆在实际使用中的性能,为车企提供更有针对性的改进建议。With the rapid development of the new energy vehicle industry,in order to improve vehicle reliability and reduce vehicle energy consumption and emissions,it is very important to accurately evaluate and analyze the driving state of vehicles in practical applications.Since the construction of the vehicle condition test model is largely dependent on the actual driving state.In this paper introduces a novel approach to classify vehicle operating conditions by leveraging extensive user-generated data from public roadways,and a load spectrum acquisition scheme that conforms to the actual driving characteristics is constructed for different regions,different users and different working conditions.This method sorts out and cleans a large number of actual driving data,analyzes the dimensionality reduction and cluster analysis,and finally provides a scientific basis for the construction of reliability test conditions.The results show that the association between the user driving model and the vehicle reliability test model is helpful to verify the performance of the vehicle in actual use and provide more targeted improvement suggestions for car companies.

关 键 词：车辆可靠性测试 公共道路大数据 主成分分析 K-MEANS聚类 

分 类 号：U41[交通运输工程—道路与铁道工程]