一种适合城市混合道路行驶工况的构建方法

A Construction Method Suitable for Driving Cycles of Urban Mixed Roads

针对城市规模和路网扩张带来的城市道路新特点,利用10辆城市乘用车按各自目的地跟随车流自主驾驶,实时采集车辆行驶数据,其路径覆盖城区、环城快速路、环线高速的城市混合道路。基于聚类法和马尔科夫模型,提出一种适合城市混合道路的工况构建方法。运用主成分分析实现特征参数的降维,以Silhouette值确定K值,利用K均值聚类法将2869个运动学片段分为3类;再将每类行驶工况抽象为一个速度随时间变化的马尔科夫过程,划分成加速、减速、巡航及怠速4种工况并计算其转移概率;根据转移矩阵合成候选工况,以CPV值筛选出目标工况,得到目标工况的整体特征参数平均误差为4.44%。结果表明,所提构建方法能较好适应城市混合道路综合行驶工况的构建。

In view of the new characteristics of urban roads brought about by urban scale and road network expansion,10 urban passenger cars are used to drive autonomously with traffic flow according to their respective destinations,and cars driving data are collected in real time.The routes cover urban mixed roads with urban area,ringcity expressway and ringline expressway.Based on the clustering method and Markov model,a construction method suitable for urban mixed roads is proposed.The principal component analysis is used to reduce the dimension of feature parameters,K value is determined by Silhouette value,and 2869 kinematic segments are classified into three categories by K-means clustering method.Each driving cycle is abstracted as a Markov process with time-varying speed,which is divided into four driving cycles:acceleration,deceleration,cruise and idling,and their transfer probability is calculated.Candidate driving cycle is synthesized based on the transfer matrix,the target driving cycle is selected by the CPV value,and the average error of the overall characteristic parameters of the target driving cycle is 4.44%.The results show that the proposed method can well adapt to the construction of comprehensive driving cycle of urban mixed roads.