China Safety Science Journal ›› 2017, Vol. 27 ›› Issue (2): 98-103.doi: 10.16265/j.cnki.issn1003-3033.2017.02.018

• Safety Science of Engineering and Technology • Previous Articles     Next Articles

Analyzing traffic crash under iced and snow weather condition based on Adaptive Lasso and RF

ZHAO Wei1,2,3, XU Liangjie1, RAN Bin3,4, WANG Jizhou1, ZHANG Xuan3   

  1. 1 School of Transportation, Wuhan University of Technology, Wuhan Hubei 430063, China
    2 School of Economics and Management, Inner Mongolia University of Science & Technology, Baotou Inner Mongolia 014010, China
    3 Civil Engineering, University of Wisconsin, Madison 53705, America
    4 School of Transportation, Southeast University, Jiangsu Nanjing 210096, China
  • Received:2016-11-01 Revised:2017-01-11 Online:2017-02-28 Published:2020-11-22

Abstract: In order to analyze the factors affecting traffic crashes under ice and snow weather conditions and make clear of the specific impact factors by driving environment, driver and vehicle conditions. This study analyzes risk factors which are differently relative to crash severity based on hybrid algorithm of Adaptive Lasso and RF which was trained with ten-year three hundred thousand crash data. The results show that the hybrid algorithm is superior to the support vector machine(SVM), the classification and regression trees(CART) and RF in both the accuracy and the fitting degree,that there is the most significant correlation between environmental factors and crashes,that accidents on slopes,curved roads,and at intersections are caused by the decrease in the road surface skid resistance,that there is a negative correlation between fatal crashes and use of safety belts,that woman is liable to make mistakes in driving,and that local drivers usually overestimate their driving skills, which leading to crashes.

Key words: highway, traffic accident, Adaptive Lasso, random forest(RF), iced and snow, big data analysis

CLC Number: