CAJ | 학술논문

本文中采用主成分分析和模糊聚类相结合的行驶工况识别方法进行纯电动汽车续驶里程的估算。首先选取20个具有代表性的循环工况数据,将其划分为215个工况片段,并选用12个特征参数对其进行主成分分析、模糊C聚类分析和行驶工况识别；然后在MATLAB/Simulink下建立纯电动汽车整车模型,进行行驶工况识别、整车能量消耗和续驶里程仿真估算；最后在转鼓试验台上进行ECE15工况下实车测试验证,结果表明：续驶里程仿真估算值与测试值的最大绝对误差为1.905km,平均绝对误差为0.742km,相对误差小于3%。
본문중채용주성분분석화모호취류상결합적행사공황식별방법진행순전동기차속사리정적고산。수선선취20개구유대표성적순배공황수거,장기화분위215개공황편단,병선용12개특정삼수대기진행주성분분석、모호C취류분석화행사공황식별；연후재MATLAB/Simulink하건립순전동기차정차모형,진행행사공황식별、정차능량소모화속사리정방진고산；최후재전고시험태상진행ECE15공황하실차측시험증,결과표명：속사리정방진고산치여측시치적최대절대오차위1.905km,평균절대오차위0.742km,상대오차소우3%。
In this paper, a driving cycle identification method is adopted, which combines principal compo-nent analysis with fuzzy clustering, to estimate the driving range of battery electric vehicle. Firstly twenty represent-ative driving cycle data are selected and divided into 215 cycle segments, and 12 characteristic parameters are cho-sen to conduct principal component analysis, fuzzy C-means clustering and driving cycle identification. Then a mod-el for battery electric vehicle is established with MATLAB/Simulink to perform driving cycle identification and the simulation estimations of vehicle energy consumption and driving range. Finally a real vehicle validation test is car-ried out on drum test bench with ECE15 cycle. The results show that compared with test data, the maximum abso-lute error of simulated estimates is 1. 905km, and the corresponding average absolute error and relative error are 0. 742km and less than 3% respectively.