因子分析在汽车行驶工况构建过程中的应用

Application of Factor Analysis in Driving Cycle Development

针对汽车行驶工况构建过程,提出了因子分析结合模糊聚类的工况构建方法,并利用该方法建立了城市道路代表性行驶工况。采用因子分析法对大量运动学片段的特征参数进行压缩,选出能代表原始片段大部分特征信息的3个公因子来进行模糊聚类分析;由聚类分析的结果,根据相关系数理论从每一类片段中提取代表性片段来构建典型行驶工况。合肥市行驶工况试验数据分析结果表明:典型行驶工况的速度-加速度矩阵与试验数据的平均误差为13.01%,构建工况能综合反映合肥市道路的交通状况。

As for the reseach of construction process of vehicle driving cycle,a method of factor analysis combined with fuzzy clustering was proposed herein,and a representative driving cycle was developed urban roads by this method.The method of factor analysis was applied to reduce the characteristic parameters of many microtrips,all microtrips were classified by using 3 common factors that can represent most of the features of the original microtrips;Based on the results of cluster analysis and the correlation coefficient theory,some representive microtrips were extracted from each type of microtrips to construct a typical driving cycle.The application of the driving cycle model to the roads in Hefei city shows that the average error in obtained typical driving cycles is 13.01% compared with the experimental data in terms of V-A matrix of the typical driving cycle.So the typical driving cycle can reflect Hefei road traffic condition comprehensively.