基于路测数据的汽车行驶工况构建方法研究  

Research on construction method of vehicle driving condition based on road test data

作　　者：阎柄辰 Yan Bingchen(Zibo City Environmental Pollution Control Center,Zibo 255025,China)

机构地区：[1]淄博市环境污染防控中心,山东淄博255025

出　　处：《汽车知识》2024年第1期110-112,共3页Auto Know

基　　金：国家自然科学基金(51508315)。

摘　　要：实现车辆性能精确标定和能耗、排放认证,构建符合实际的行驶工况是基础性工作。本文以道路实验数据为基础,经过数据预处理以及运动学片段的分割,利用主成分分析法确定工况建立的主成分,再基于SOM神经网络+K-means聚类将运动学片段聚合成低速、中速和高速3大类工况,根据各类工况特征选取代表性的运动学片段合成行驶工况。将合成的行驶工况与实验采集数据进行对比,发现特征值的相对误差在10%以内,说明所合成的工况能够反映轻型车的实际道路行驶状况。In order to achieve accurate vehicle performance calibration and energy consumption and emission certification,it is a basic work to build a driving condition that conforms to the reality.Based on the road experiment data,this paper uses the principal component analysis method to determine the principal components established by the driving conditions through data preprocessing and the segmentation of kinematic segments.Then,based on SOM neural network+K-means clustering,the kinematic segments are aggregated into three types of driving conditions:low speed,medium speed and high speed,and representative kinematic segments are selected to synthesize driving conditions according to the characteristics of various working conditions.By comparing the synthesized driving conditions with the experimental data,it is found that the relative error of the characteristic values is less than 10%,which indicates that the synthesized driving conditions can reflect the actual road driving conditions of light vehicles.

关 键 词：行驶工况 运动学片段 主成分分析 SOM神经网络 K-MEANS聚类 

分 类 号：U461[机械工程—车辆工程]