针对实际交通系统时变复杂的特征和交通流变化的不确定性,应用小波分析理论,对原始交通数据进行了消噪处理,使消噪后的数据更能反映交通流的本质及变化规律;采用综合自回归移动平均(ARIMA)时间序列模型对交通流进行预测;并对实测交通数据进行验证分析.结果表明,该方法具有较高的预测精度,可用于交通流的实时动态预测.
Based on the analysis of the fundamental characteristics of the field traffic system and traffic flow, this paper presents a new method of traffic flow prediction. The theory of wavelet analysis is applied to eliminating the noise from the real-time traffic data to better reflect a real traffic flow. Based on the processed traffic data, the time-series model, autoregressive integrated moving average model (ARIMA) is used to predict traffic flows in different periods. Finally, the paper presents numerical examples on the field data to testify the the proposed model.