基于灰色Verhulst和HMM的交通事故预测算法

Traffic Accident Prediction Algorithm Based on Grey Verhulst and HMM

道路交通事故预测是保证道路佳通安全的重要技术,以往的道路交通事故预测方法往往具有预测精度不高和收敛速度慢的缺点,为此,设计了一种基于灰色Verhulst模型和隐形马尔科夫链的交通事故预测方法;首先,采用灰色Verhulst模型对观察事件在下一时刻的状态进行预测,采用最小二乘估计法去估计模型中的参数,将预测结果用于初始化HMM模型,并采用前向算法和后向算法对HMM模型进行训练,获得最终的初始分布矩阵、状态转移概率和观察概率分布矩阵,然后采用最终的HMM模型进行交通事故预测;仿真试验结果表明:文中方法能有效地实现交通事故预测,较其它方法相比,具有预测精度高和收敛速度快的优点,具有一定的优越性。

Road traffic accident prediction is an important technology to guarantee the safety of road,the given road traffic accident prediction method has the low prediction accuracy and low convergence speed,therefore,a prediction method based on grey Verhulst model and HMM is proposed.Firstly,the grey Verhulst model is used to predict the state of the next time,the least square method is used estimate the parameters of the model,the predicting result is used to initialize the HMM model,and the forward and backward algorithm are designed to train the HMM model,and the final initial distribution matrix,state transferring probability and observing probability distribution matrix,and the final HMM model can be used to predict the traffic accident.The simulation result shows the method in this paper can effectively realize the traffic accident prediction,and compared with the other methods,it has the high predicting accuracy and quick convergence speed.Therefore,it has big priority.