论文部分内容阅读
交通崩溃事件会造成道路通行能力下降,成为导致城市快速路拥堵的主要原因之一,精准的短时交通崩溃事件预测在交通管理与控制中具有重要意义。该文以美国加州高速公路性能评估系统(PeMS)提供的交通流数据为基础,对道路的崩溃状态进行了分级定义,并以道路崩溃状态为隐状态、道路占有率为显状态,结合二者之间的联系,建立了隐Markov模型。通过数理统计,对模型参数进行了学习,最后采用Viterbi算法对该模型进行了求解,实现了快速路交通崩溃事件的预测。预测结果与实际数据的对比表明:该方法能预测大部分的交通崩溃事件。
The traffic collapse event will cause the capacity of the road to pass down and become one of the main causes of congestion in the expressway. The accurate prediction of short-time traffic collapse event is of great significance in traffic management and control. Based on the traffic flow data provided by the California Highway Performance Evaluation System (PeMS), this paper classifies the collapse status of a road and defines the road collapse state as a hidden state and the road share as a dominant state. Combining the two The relationship between the established hidden Markov model. Through the mathematical statistics, the parameters of the model were studied. Finally, the Viterbi algorithm was used to solve the model and the prediction of the traffic collapse event of expressway was realized. A comparison of the forecast results with the actual data shows that the method can predict most of the traffic collapse events.