A method for predicting traffic flow using a reinforced long short-term memory model
By combining a reinforced long short-term memory model with convolutional neural networks and attention mechanisms, the problem of insufficient spatiotemporal feature capture in traditional transportation systems is solved, achieving higher accuracy in traffic flow prediction and signal control, and improving the overall performance of the transportation system.
CN122416731APending Publication Date: 2026-07-17HENAN POLICE ACAD
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- HENAN POLICE ACAD
- Filing Date
- 2026-05-06
- Publication Date
- 2026-07-17
Smart Images
Figure CN122416731A_ABST
Abstract
本发明公开了一种利用强化型长短时记忆模型预测交通流量的方法，涉及交通控制技术领域，包括以下步骤：S1、基于强化LSTM的交通流量预测模型构建；S2、基于云平台路网交通流平衡的信号控制。本发明采用上述一种利用强化型长短时记忆模型预测交通流量的方法，充分考虑路网交通流量的空间特征和路网交通流量的时间特征，提高交通流量的预测精度，减小误差；基于云平台路网交通流平衡的信号控制将局部信号自适应控制与区域信号协同控制结合，考虑路网交通流平衡的信号控制分层协同策略，有效降低交叉口平均延误时间。
Need to check novelty before this filing date? Find Prior Art