摘要
根据当前交通流预测模型发展的情况,提出采用卡尔曼滤波模型、指数平滑模型和非参数回归模型进行预测融合的方法,该方法在提高系统预测速度的基础上也能提高预测精度,因此,具有很好的推广价值。
According to the development condition of the traffic flow forecasting's models, the comprehensive prediction method, which involve kalman filtering model, exponential smoothing model and nonparametric regressive model. Is put forward. The model not only improve system's prediction speed, but also improve system's prediction precision, so, it has good value to use more.
出处
《交通标准化》
2007年第4期207-210,共4页
Communications Standardization
关键词
短时交通流
交通流预测
模型
信息融合
short-time traffic volume
traffic flow forecasting
models
data fusion