基于灰色马尔可夫的道路交通事故预测

探讨灰色马尔可夫模型在道路交通事故中的具体应用。灰色模型适用于短期、数据量少和波动不大的预测问题,在长期预测时,数据序列拟合较差,预测精度偏低;而马尔可夫链适用于长期、数据序列随机波动大的预测问题。灰色马尔可夫模型结合了灰色GM(1,1)模型和马尔可夫理论的优点,利用灰色模型进行长期预测,再利用马尔可夫链理论进行波动状态预测,最后得到期望值。该模型克服了随机波动性数据对道路交通事故预测精度的影响,提高了灰色预测的准确度。实例结果,证明灰色马尔可夫GM(1,1)模型具有较好的应用价值,为道路交通安全管理提供了有用依据。

Road Traffic Accident Forecast Based on Gray-Markov Model

In the light of the features of traffic accidents, this paper makes a study of the concrete application of Gray-Markov model in road traffic accidents. In view that gray theory prediction is suitable for this kind of system with the characteristics of few data, short-time running and little fluctuation and that Markov chain theory is suitable for long-term prediction with big fluctuation, Gray-Markov model is presented by combining the advantages of both gray prediction and Markov prediction chain theory. This new model firstly makes a long-term prediction through Grey model, then acquires the range of fluctuating value through Markov model, and finally gains the expectation. This model avoids the effect of random fluctuation data on the prediction precision of road traffic accidents and improves the accuracy of grey forecast. The exemplification of this model in the forecast of domestic road traffic accidents shows that this model possesses good practicability and could provide valuable reference for road safety management.