摘要

Principal component analysis (PCA) is adopted to study the main constitution structures of driver's operation modes and a PCA algorithm created with Matlab is used to analyze the data obtained by questionnaire, from which three sets of driver behavior data of overtaking, steering and cutting-in maneuver are in particular selected for analysis to obtain the loading matrix and contribution rate of PCA corresponding to those three operation modes. The results show that through the study on driver's behaviors and corresponding operation modes, the driving operation actions having more significant effects on specific driving behaviors can be found and hence the identification of personalized driving behaviors can be achieved inversely based on the relationship between them, providing a technique basis for subsequent design of advanced driver assistant system.

全文