Identification and Difference Analysis of Motorcycle Drivers' Driving Style Based on Natural Driving Data

In order to deeply understand the driving behavior characteristics of different motorcycle drivers, based on the natural driving data of Shaanxi Chebaidu Motorcycle Company, five types of bad driving behaviors are used as driving behaviors: sharp acceleration, sharp deceleration, sharp right turn, left sharp turn and speeding tendency. The style characteristic index, based on the SOM-K-Means clustering calculation, divides the driving style of drivers into three categories: cautious, mild, and aggressive, and the corresponding driver proportions are 88%, 10%, and 1% of the sample drivers, respectively. The statistical analysis of drivers with different driving styles after clustering shows that the SOM-K-Means algorithm can effectively realize the clustering of drivers with different driving styles.