The invention relates to a short-time traffic flow prediction method and system, and relates to the field of traffic flow prediction, and the method comprises the steps: obtaining traffic flow data at a current moment; inputting the traffic flow data at the current moment into a Bi-LSTM model of a traffic flow prediction model for bidirectional time sequence feature extraction to obtain traffic flow bidirectional time sequence features at the current moment; the traffic flow prediction model comprises a Bi-LSTM model, an attention mechanism and an activation function which are connected in sequence; inputting the traffic flow bidirectional time sequence characteristic at the current moment into the attention mechanism for processing to obtain a traffic volume time sequence characteristic value at the current moment; and activating the traffic volume time sequence characteristic value at the current moment by using the activation function to obtain the traffic volume at the next moment. According to the invention, the prediction precision of the short-time traffic flow can be improved.
本发明涉及一种短时交通流预测方法及系统,涉及交通流预测领域,方法包括:获取当前时刻的交通流数据;将所述当前时刻的交通流数据输入交通流预测模型的Bi‑LSTM模型进行双向时序特征提取,得到当前时刻的交通流双向时序特征;所述交通流预测模型包括依次连接的Bi‑LSTM模型、注意力机制和激活函数;将所述当前时刻的交通流双向时序特征输入所述注意力机制进行处理,得到当前时刻的交通量时序特征值;将所述当前时刻的交通量时序特征值利用所述激活函数进行激活,得到下一时刻的交通量。本发明能够提高对短时交通流的预测精度。
Short-term traffic flow prediction method and system
一种短时交通流预测方法及系统
2022-07-12
Patent
Electronic Resource
Chinese
Urban short-term traffic flow prediction method, system and equipment
European Patent Office | 2024
|Short-Term Traffic Flow Prediction: Using LSTM
IEEE | 2020
|