一种面向人群密集情态下的缓冲区域智能调控方法
By combining pedestrian dynamics simulation and a deep dual-Q network-based intelligent control method, the adaptive control problem of buffer fence facilities under dynamic crowd flow was solved, achieving safe and stable congestion management and efficient traffic flow.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- QINGDAO UNIV OF TECH
- Filing Date
- 2026-06-22
- Publication Date
- 2026-07-17
AI Technical Summary
Existing buffer barriers are unable to adapt and dynamically adjust to dynamic crowd flow, resulting in low traffic efficiency and the risk of congestion and displacement, and are unable to effectively cope with dynamic changes in passenger flow.
Pedestrian dynamics simulation software is used to model passenger flow. Combined with data-driven system identification and model predictive control, a state-space predictive dynamic identification model (SSM) is constructed. Combined with deep double-Q network reinforcement learning, intelligent regulation is achieved. The control strategy is optimized through state estimator and multi-objective cost function. Future prediction terms and soft constraints are introduced to carry out adaptive facility adjustment.
It achieves adaptive dynamic control under safety constraints, effectively suppresses congestion transfer, improves traffic efficiency and reduces facility switching frequency, reduces disturbance to passenger behavior, and improves the deployment efficiency of the algorithm in real environment.
Smart Images
Figure CN122414760A_ABST