基于改进的灰色马尔可夫链模型的交通事故预测

Traffic Accident Prediction Based on Improved Grey-Markov Chain Model

交通事故预测是交通安全评价、规划和决策的基础.灰色预测适合于数据量少和波动小的系统对象,而马尔可夫链理论适用于预测随机波动大的动态过程.为克服一般灰色马尔可夫链模型运用的转移概率矩阵固定不变而影响预测精度的问题,本文建立了改进的灰色马尔可夫链模型.采用滑动转移概率矩阵方法,去掉最老数据并补充最新数据,从而建立新的一步转移概率矩阵.借助改进的灰色马尔可夫链模型,对全国2002-2004年交通事故10万人口死亡率进行了预测分析.结果表明,改进的灰色马尔可夫链模型比一般灰色马尔可夫链模型的预测范围更准确,预测精度更高.

The prediction of traffic accident is the basis of transportation safety,assessment and decision-making.Grey prediction is suitable for such kinds of system objects with short data and little fluctuation,as well as Markov chain theory fits for forecasting dynamic process with random and large fluctuation.To overcome the problem about prediction accuracy is affected during the general grey-Markov chain model for using fixed transition probability matrix;this paper sets up an improved grey-Markov chain model.It uses sliding transition probability matrix,removing the oldest data and adding the latest data to create a new step transition probabiUty matrix.By means of the improved grey-Markov chain model,the100,000 population death rate of traffic accident from 2002 to 2004 has been predicted.The result shows that-the prediction accuracy of improved gray-Markov chain model is better than that of the ordinary grey-Markov chain model.