Abstract: In this paper, we use wavelet transfonn to decompose taffic data into individual components and perfom the prediction of each component by ARMA model, we obtain traffic prediction by each component' s prediction. Experimental results shows that the menthod has high precision .

Key words: wavelet analysis, ARMA model, short-term taffic prediction

摘要： 本文利用小波分析方法对交通流数据进行不同尺度的分解，并对不同尺度的分量建立ARMA预测模型，再由 各尺度的预测得到交通流的预测．实验显示，本方法具有较好的预测精度

关键词: 小波分析, ARMA模型, 短时交通流预测