基于SARIMA模型的城市道路短期交通流预测研究

以城市道路短期交通特征为基础，提出在使用季节自回归求和移动平均（SARIMA）模型进行预测时还应该考虑交通流日周期性特征，以提高模型的预测能力。预测研究不仅考虑了模型的样本内预测而且考虑了样本外预测，并通过两个预测误差统计量，即预测误差绝对值均值和预测误差百分比绝对值均值，来对模型预测表现进行衡量与比较。研究结果显示，虽然只考虑城市道路短期交通流数据中周周期性的SARIMA模型预测能力比只考虑日周期性特征的SARIMA模型好，但是却比既考虑交通流数据中周周期性又考虑数据中的日周期性的SARIMA模型表现差。

Study on Urban Road Short-term Traffic Flow Forecasting Based on SARIMA Time Series Model

Daily pattern should be considered during forecasting with the seasonal autoregressive integrated moving average(SARIMA) models based on the characteristic of urban road short-term traffic flow.The forecasting experiment consists of in-sample and out-of-sample forecasting.Two forecast error statistics,namely,the mean absolute error(MAE) and the mean absolute percentage error(MAPE),were employed to measure forecasting ability of models.The results show that the forecasting performance of the SARIMA model only considering weekly pattern is better than that of the SARIMA model only considering daily pattern,but worse than that of the SARIMA model considering both weekly pattern and daily pattern.