交通事故预测方法研究

交通事故预测是在对已发生事故的数据进行统计、分析和处理的基础上,以事故发生原因和发生的规律为依据,对目前尚未发生的事故预先做出的推测、判断。介绍了几种主要的公路交通事故预测方法,包括线性回归预测法、非线性回归预测法、时间序列预测法的移动平均法与指数平滑预测法、灰色马尔科夫链预测法、基于车速的交通事故贝叶斯预测法、道路交通事故的灰关联分析及神经网络预测方法等,国内、外传统的和主流的交通事故预测的方法,并对几种主要的公路交通事故预测方法的适用条件、优缺点进行了较详细的比较,为相关部门选择合适的交通事故预测方法提供参考。

Reaserch of Traffic Accident Prediction Method

Predicting the traffic accidents is making prediction based on the statistics and analysis of happened accidents data,depending on the reason and regulation of accidents.The article gets on the traditional and prediction methods such as Linear regression analysis,Non-linear regression analysis,Moving Average and Exponential Smoothing of Time series,Grey Markov Model for Forecasting the Death Toll of Road Accidents,Bayesian Prediction of Traffic Accident Based on Vehicle Speed,Grey correlative Analysis and Neural Network of Traffic Accident Prediction Method and so on..Besides that,by comparing of the applying conditions,the advantages and disadvantages of main prediction methods,it provides proper traffic accident prediction methods to related department.