The invention discloses a reinforcement learning-based ramp signal control optimization method and system; the method comprises the steps: a ramp intersection optimization control step: detecting a real-time traffic flow on a road through single-point adaptive control, selecting a ramp signal control scheme by an upper computer according to the real-time traffic flow, and building an SARSA signal control model; a model parameter calibration step: acquiring a vehicle following and lane changing model, and calibrating parameters of the vehicle following and lane changing model; and a simulation step: training the SARSA signal control model and the vehicle following and lane changing model after parameter calibration according to a preset demand to obtain an optimized ramp signal control scheme. According to the method, the reinforcement learning-based on-ramp signal control optimization method on the expressway is designed, and on the basis of ramp control method verification and effect evaluation of traffic simulation, the SUMO simulation is used for verifying the effect, so that a new thought and a new method are provided for later theoretical research and engineering application.
本发明公开了一种基于强化学习的匝道信号控制优化方法和系统,包括:匝道交叉口优化控制步骤,通过单点自适应控制检测道路上的实时交通流,上位机根据所述实时交通流选择匝道信号控制方案,建立SARSA信号控制模型;模型参数标定步骤,获取车辆跟驰与换道模型,对所述车辆跟驰与换道模型的参数进行标定;仿真步骤,根据预设的需求训练所述SARSA信号控制模型和标定参数后的车辆跟驰与换道模型,得到优化的匝道信号控制方案。本发明通过设计基于强化学习的快速路上匝道信号控制优化方法,并基于交通仿真的匝道控制方法验证及效果评价,使用SUMO仿真验证效果,为以后的理论研究和工程应用提供新的思路和方法。
Ramp signal control optimization method and system based on reinforcement learning
一种基于强化学习的匝道信号控制优化方法和系统
2021-09-17
Patent
Elektronische Ressource
Chinesisch
Ramp control method based on offline reinforcement learning and macroscopic model
Europäisches Patentamt | 2022
|Multi-ramp coordination control method based on different strategy reinforcement learning
Europäisches Patentamt | 2022
|Vehicle ramp entrance confluence control method based on deep reinforcement learning
Europäisches Patentamt | 2023
|Reinforcement Learning for Ramp Control: An Analysis of Learning Parameters
DOAJ | 2016
|