混合动力汽车驾驶风格识别的研究

摘要

鉴于驾驶员驾驶风格对混合动力汽车燃油经济性和排放性能有重要影响,本文中旨在通过对驾驶风格进行分类和识别,提高整车控制策略对驾驶风格的适应性,以改善整车的燃油经济性.首先以某款混合动力汽车为对象,对不同驾驶员的驾驶风格进行了实车道路试验和数据采集,接着采用主成分分析提取出表征驾驶风格的综合特征参数,并应用K-均值聚类对驾驶风格进行了聚类分析,最后在此基础上利用支持向量机算法对驾驶风格进行了识别.结果表明,驾驶风格的识别精度达到了90%以上,为混合动力汽车能量管理策略的进一步自适应优化奠定了基础.%In view of the significant effects of driver's driving style on fuel economy and emission perform-ance of hybrid electric vehicles, this paper aims to enhance the adaptability of vehicle control strategy to driver's driving style through the classification and identification of drivers' driving styles for improving the fuel economy of vehicle. Firstly,with a hybrid electric vehicle as object,a real vehicle test with data acquisition is conducted by dif-ferent drivers with different driving styles.Then principal component analysis is adopted to extract the comprehensive characteristic parameters of driving styles,and K-means clustering is applied to the cluster analysis of driving styles. Finally,on these bases,support vector machine algorithm is used to identify driving styles.The results show that the recognition accuracy of driving styles reaches more than 90%,laying a solid foundation for the subsequent adaptive optimization of energy management strategy for hybrid electric vehicles.