This study utilizes data obtained from vehicle positioning, speed, and engine status through satellite positioning system monitoring terminals. The data is processed using spherical interpolation of latitude and longitude to handle missing values. The study establishes indicators for identifying bad driving behaviors, including speeding, rapid acceleration and deceleration, prolonged vehicle idling, and fatigue driving. The traditional DBSCAN algorithm is improved by incorporating both temporal and spatial search ranges into the algorithm, enabling cluster analysis of bad driving behaviors from both temporal and geographical perspectives. Finally, the Graham algorithm is utilized to calculate cluster boundaries, determining the specific locations and boundaries of high incidence areas of bad driving behaviors. This research aids regulatory authorities in more precisely supervising risk areas during transportation processes in terms of time and location.
本站僅提供期刊文獻檢索。
