Time-space Characteristics Distribution of Traffic Accidents Based on Clustering Analysis

摘要

This paper uses clustering analysis method to explore time-space characteristics distribution of traffic accidents, losses converted into cash, number of deaths and injuries. We select hierarchical clustering method to analyze about 458 thousands road traffic accidents in china from 2009 to 2010. And the time-space characteristics distributions include month distribution, annual average daily distribution, annual average daily time of occurrence distribution, road type distribution, road alignment distribution. The results show that traffic accidents on expressway are bound up with these characteristics.