论文总字数：28126字

摘 要

论文作者签名：＿＿＿＿＿导师签名：＿＿＿＿日期：＿＿＿＿

交通要素与交通事故的关联分析

21113231 潘东昊

指导教师 陈淑燕

摘 要

本文采用了主成分分析法和Logistic回归分析方法将交通要素与交通事故之间进行了关联分析。本次研究所作关联分析在交通事故预防方面以及道路交通管理整治方面能起到积极作用。

本文首先对美国加州I-880N高速公路进行了数据预处理，对2008.01~2010.12的事故数据、交通流数据、道路几何特征数据、天气数据进行了匹配和筛选，删除了冗余数据。其次，本文对预处理之后的数据进行主成分分析以达到降维的目的。在数据降维之后，采取了Logistic回归分析对七个主成分进行回归分析，剔除没有统计学意义的两个主成分。最后，将主成分分析的结果和Logistic回归分析的结果相结合，分析了不同类型的交通事故与交通要素之间的关联。

研究成果表明，道路拥堵程度越大、流量越大、车道数越少，则发生危险事故的概率较之发生交通流失效的概率越大；同理，道路拥挤程度越小、流量越大、路肩宽度和路中分带宽度越大，则发生危险事故的概率较之发生普通交通事故的概率越大。并且，根据分析给出了各交通要素增加时，发生特定事故类型的概率变化。研究成果有助于有针对性地改善道路交通条件，有助于减少普通交通事故和危险交通事故的发生。

关键词：交通事故；交通要素；数据匹配；主成分分析；Logistic回归分析

An Analysis of the Correlation between Traffic Factors and Traffic Accidents

21113231 PanDonghao

Supervised by ChenShuyan

Abstract

In this paper, principal component analysis and Logistic regression analysis are used to analyze the relationship between traffic factors and traffic accidents. The association analysis can play a positive role in traffic accident prevention and road traffic management.

In this paper, the data preprocessing of the I-880N expressway in California is carried out, and the accident data, traffic flow data, road geometric characteristic data and weather data are matched and screened from 2008.01 to 2010.12, and redundant data are deleted. Secondly, this paper makes a principal component analysis on the data after pretreatment to achieve the purpose of dimension reduction. After the data was dimensioned, Logistic regression analysis was used to analyze the seven principal components, and the two principal components were not statistically significant. Finally, the results of principal component analysis and Logistic regression analysis are combined to analyze the association between different types of traffic accidents and traffic factors.

The results show that the greater the degree of road congestion, the greater the traffic, the smaller the number of lanes, the probability of occurrence of dangerous accidents is greater than the probability of occurrence of traffic failure; Similarly, the greater the degree of road congestion, the greater the flow, The greater the probability of occurrence of a dangerous accident in the road is greater than the probability of an ordinary traffic accident and the probability of a specific type of accident occurs when the traffic factor is increased, which is helpful to improve road traffic conditions, and to reduce the incidence of ordinary traffic accidents and dangerous traffic accidents.

KEY WORDS: traffic accident; traffic factor; data match; principal component analysis; Logistic regression analysis

目 录

摘 要 I

Abstract II

第一章 绪 论 1

1.1 引言 1

1.2研究现状 1

1.3 关联分析思路 2

1.4 论文结构 3

第二章 数据来源及数据匹配 4

2.1 数据描述 4

2.1.1 数据来源 4

2.1.2 数据采集流程 5

2.2 数据匹配 5

2.2.1 事故数据和道路几何特征数据的匹配 5

2.2.2 事故数据与交通流数据之间的匹配 6

第三章 主成分分析和Logistic回归分析原理 7

3.1 主成分分析 7

3.2 Logistic回归分析 8

3.2.1Logistic回归分析介绍 8

3.2.2Logistic回归系数的解释 9

第四章 实例分析 12

4.1 数据降维 12

4.2 主成分与事故类型的回归分析结果 15

第五章 总 结 23

参考文献（References） 24

致 谢 25

绪 论

本章介绍了本次研究的背景和意义，分析了目前的研究现状，阐述了本次研究的具体思路并且介绍了本文的总体结构框架。

剩余内容已隐藏，请支付后下载全文，论文总字数：28126字