【摘要】：用南京地區(qū)2012年逐日交通事故數(shù)據(jù)和實測氣象資料,在考慮自相關(guān)性的前提下,通過多因子時間序列分析,構(gòu)建南京地區(qū)2012年道路日交通事故量的AR氣象影響模型,對工作日和非工作日分別分析發(fā)現(xiàn):不利氣象條件與日交通事故量,工作日比非工作日相關(guān)顯著,降水、低溫、低能見度等都與交通事故顯著相關(guān),中雨以下日降水量與日交通事故量呈正相關(guān),日平均氣溫在2~12℃、日最低能見度在200~500 m范圍內(nèi),都與日交通事故量呈顯著負相關(guān);但大的降水、極端氣溫、低能見度與發(fā)生交通事故的相關(guān)性反而減小。又根據(jù)不同氣象要素在日交通事故量中的作用大小,構(gòu)建氣象影響逐步線性回歸模型。最后,比較兩種模型的優(yōu)劣,擬合優(yōu)度分析顯示,工作日期間AR模型的擬合效果比逐步回歸模型更好。
[Abstract]:Based on the daily traffic accident data and measured meteorological data in Nanjing in 2012, and considering the autocorrelation, the AR meteorological influence model of daily road traffic accidents in Nanjing in 2012 is constructed by multi-factor time series analysis. The analysis of working days and non-working days shows that the unfavorable meteorological conditions are significantly correlated with the daily traffic accidents, and the correlation between working days is more significant than that of non-working days. Precipitation, low temperature and low visibility are all significantly correlated with traffic accidents. The daily precipitation below moderate rain was positively correlated with the daily traffic accident volume. The daily average temperature was 2 擄12 鈩，

本文編號：2481451