The invention provides a road intersection traffic signal control method and equipment. The method provided by the invention comprises the following steps: in response to application of a traffic signal phase to the road intersection, obtaining a road vehicle state of the road intersection; providing the road vehicle state to a Direct Light reinforcement learning model to obtain a traffic signal phase duration parameter output by the Direct Light reinforcement learning model, and multiplying the traffic signal phase duration parameter output by the Direct Light reinforcement learning model by the reference duration of the traffic signal phase to obtain the traffic signal phase. Obtaining the estimated traffic signal phase duration of the traffic signal phase; and updating the traffic signal phase applied to the road intersection after the traffic signal phase duration of the traffic signal phase has elapsed.
提供了道路交叉口交通信号控制方法及设备。所提供的方法包括,响应于向所述道路交叉口施加了交通信号相位,获取所述道路交叉口的道路车辆状态;将所述道路车辆状态提供给DirecLight强化学习模型,获得所述DirecLight强化学习模型输出的交通信号相位持续时间参数,用所述DirecLight强化学习模型输出的交通信号相位持续时间参数与所述交通信号相位的参考持续时间相乘,得到估计的所述交通信号相位的交通信号相位持续时间;在时间经过了所述交通信号相位的交通信号相位持续时间后,更新施加给所述道路交叉口的交通信号相位。
DirectLight: estimation of signal intersection real-time traffic signal phase duration based on reinforcement learning
DirectLight:基于强化学习估计信号交叉口实时交通信号相位持续时间
2024-02-06
Patent
Elektronische Ressource
Chinesisch
IPC: | G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS |
Real-time dynamic allocation method and system for intersection traffic signal duration
Europäisches Patentamt | 2021
|Multi-intersection traffic signal control method based on deep reinforcement learning
Europäisches Patentamt | 2024
|Multi-intersection traffic signal control method based on deep reinforcement learning
Europäisches Patentamt | 2023
|Traffic intersection unmanned signal control real-time optimization method
Europäisches Patentamt | 2021
|