河北省道路交通事故灰色预测模型的构建与应用

发布时间：2018-11-08 12:56

【摘要】：道路交通事故预测与预防是减少交通事故，提高道理交通安全水平的主要手段之一，逐渐发展成为道路交通安全研究的热点问题。道路交通系统是一个动态的不确定的系统，交通事故的发生具有随机性。针对道路交通系统“部分信息已知，部分信息未知”的特性，采用灰色系统理论，着重研究“小样本、贫信息”的不确定问题及“外延明确、内涵不明确”的随机对象，实现“少数据建模”。本文着重研究道路交通事故宏观影响因素对死亡人数的影响，利用灰色关联分析方法确定主要影响因子，建立预测指标体系。通过构建基于灰色关联分析的多因子灰色预测模型MGM(1, N)，利用灰色生成算子的作用弱化随机性，分析道路交通系统内在联系，挖掘潜在规律，经过灰色灰色差分方程与灰色微分方程之间的互换实现了利用离散数据序列建立连续动态的微分方程，，对河北省道路交通事故进行预测。利用河北省2000年至2011年的道路交通事故统计数据进行死亡人数的预测与精度检验。实证结果表明，基于灰色关联分析的多因子灰色预测模型是可行、有效的。相较于GM(1,1)预测模型，此方法能够更加合理、科学、准确地预测交通事故的发展态势，可以为交通相关部门提供管理与决策依据。
[Abstract]:The prediction and prevention of road traffic accidents is one of the main means to reduce traffic accidents and improve the level of traffic safety. It has gradually developed into a hot topic in the research of road traffic safety. Road traffic system is a dynamic and uncertain system, and the occurrence of traffic accidents is random. In view of the characteristic of "part information is known, part information is unknown" of road traffic system, the grey system theory is adopted to study the uncertain problem of "small sample, poor information" and the random object with "explicit extension and unclear connotation". Implement "less data modeling". This paper focuses on the study of the influence of macro factors of road traffic accidents on the number of deaths. The grey correlation analysis method is used to determine the main influencing factors and to establish a prediction index system. By constructing the grey prediction model MGM (1) based on grey relation analysis, N), weakens randomness by using the function of grey generating operator, analyzes the inherent relation of road traffic system, and excavates the latent law. Through the exchange between grey difference equation and grey differential equation, the continuous dynamic differential equation is established by using discrete data sequence, and the road traffic accident in Hebei Province is forecasted. Based on the statistics of road traffic accidents from 2000 to 2011 in Hebei Province, the death toll was predicted and the accuracy was tested. The empirical results show that the multi-factor grey prediction model based on grey correlation analysis is feasible and effective. Compared with the GM (1 / 1) prediction model, this method can predict the development situation of traffic accidents more reasonably, scientifically and accurately, and can provide management and decision-making basis for traffic related departments.
【学位授予单位】：河北科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U491.31

【参考文献】