论文部分内容阅读
针对城市道路交通流预测仅考虑正常(或没有交通事故)交通条件下短期交通流的现状,采用三制度自我激励阈值自回归(SETAR)模型进行短期交通流预测,该模型适合研究多种交通状态下交通流的动态变化行为。通过实例验证,对比分析了三制度SETAR模型和自回归求和移动平均(ARIMA)模型对城市道路5 min交通流的预测,其中后者被用作比较基准模型。研究结果表明,这个模型不仅能够合理地解释交通流的变化行为,而且比ARIMA模型在向前1步样本外预测交通流变化幅度和变化方向上均有更好的预测表现。
Aiming at the current situation of short-term traffic flow in urban road traffic flow prediction considering only normal traffic condition (or no traffic accident), the short-term traffic flow forecasting method using three-system self-motivation threshold autoregression (SETAR) model is proposed. This model is suitable for studying various traffic states Dynamic changes in traffic flow. Through case validation, the three-system SETAR model and the ARIMA model are compared to predict 5-minute traffic flow of urban roads, of which the latter is used as a comparative baseline model. The results show that this model can not only explain the change behavior of traffic flow reasonably, but also have a better prediction performance than the ARIMA model in predicting the range and direction of the change of traffic flow in the forward one-step sample.