Dynamic traffic dispersing control method and system based on traffic flow situation prediction
By constructing a dynamic effective capacity and multi-layer spatiotemporal graph convolutional neural network, combined with real-time traffic data and physical parameters, the problem of road carrying capacity prediction distortion under severe weather conditions was solved, realizing the pre-control of traffic management and ensuring the stability of urban traffic operation.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHINA NAT CHEM COMM CONSTR GRP CO LTD
- Filing Date
- 2026-06-09
- Publication Date
- 2026-07-07
AI Technical Summary
Existing traffic flow prediction and diversion technologies fail to effectively integrate the three-dimensional road geometry of cities with the physical attenuation law of underlying forces under extreme weather conditions. This results in distorted predictions of the effective dynamic traffic capacity of roads under adverse terrain and weather conditions, causing serious delays in traffic diversion and diversion instructions. This can easily lead to overflow deadlocks at urban underpasses and ground intersections.
By acquiring real-time traffic flow velocity, road surface friction coefficient, average vehicle wheelbase, and three-dimensional road longitudinal slope angle, a dynamic effective capacity is constructed. Traffic flow is predicted using a multi-layer overlay spatiotemporal graph convolutional neural network. Overflow prevention and diversion control is performed based on corrected weights and physical saturation parameters. Phase duration compensation parameters are generated to execute overflow prevention, interception, and diversion control.
It enables accurate prediction of changes in road carrying capacity under severe weather conditions, early prevention of congestion, reduction of traffic management delays, ensuring the orderly operation of core urban traffic, and avoiding traffic overflow and deadlock at traffic nodes.
Smart Images

Figure CN122347873A_ABST
Abstract
Citation Information
Patent Citations
A Multi-Time Scale Prediction Method for Road Traffic Operation Status
CN104778837B
A traffic flow prediction method
CN112927510B