Research on Driving Style Classification for Shift Schedule of Dual Clutch Transmissions

Abstract

Abstract As one of the most promising vehicle automatic transmission, the dual clutch transmissions (DCT) have become a research hotspot. In order to formulate different shift schedules of DCT to meet economic and comfort requirements, it is necessary to classify and identify driving styles based on vehicle driving data. Accurate classification of driving style is a prerequisite for effective identification, and in this research, a driving style classification method is built based on feature engineering. First, a specified road test is conducted considering the influence factors, in which the driving data is collected, and the driving style is subjectively evaluated. Subsequently, the information entropy is applied to discretize the velocity and the degree of accelerator pedal degree, where 44 feature quantities are extracted to characterize the driving style. Taking into account strong correlation and redundancy between the constructed feature quantities, the principal component analysis (PCA) is employed to reduce the dimension. Finally, the fuzzy c-means (FCM) clustering algorithm is used to classify the driving style. The successful classification rate can reach 92.16% of the subjective scoring result, and is improved by 9.81% comparing with traditional feature quantities. The results show the effectiveness of the proposed driving style classification method, which lays a foundation for the adaptive control of different driving styles for the establishment of an intelligent DCT control system.