The invention discloses a region boundary main intersection signal control method based on deep reinforcement learning. The method comprises the steps: dividing a road network into a key region and a peripheral region; building a simulation platform based on Sumo, importing the real bus travel data, setting the social traffic flow reasonably, and obtaining the MFD of a key area by means of simulation; establishing a traffic flow dynamic conservation equation between the key area and the peripheral area; on the basis of the traffic flow dynamic conservation equation established in the step S3, combining model predictive control MPC and a genetic algorithm to solve an optimal boundary control parameter; and under the condition that the optimal boundary control parameters are obtained, carrying out deep reinforcement learning to obtain an optimal region boundary main intersection signal control scheme, so the difference between the actual transfer traffic flow and the expected transfer traffic flow is minimum. According to the invention, intelligent control of boundary intersection signals of an urban road network is realized, and the in-out ratio of each area is dynamically adjusted, so that the purpose of relieving traffic jam is achieved.
本发明公开了一种基于深度强化学习的区域边界主交叉口信号控制方法,包括步骤:将路网分为关键区域与外围区域两大部分;基于Sumo搭建仿真平台,导入真实的公交出行数据,合理设置社会车流量,借助仿真获取关键区域MFD;建立关键区域与外围区域之间的交通流动态守恒方程;基于步骤S3建立的交通流动态守恒方程,结合模型预测控制MPC与遗传算法,求取最优边界控制参数;在求得最优边界控制参数的情况下,进行深度强化学习,得到最优的区域边界主交叉口信号控制方案,使得实际转移车流量与预期转移车流量之间的差距最小。本发明实现了城市路网的边界交叉口信号智能控制,动态的调节各区域间的进出比,以达到缓解交通拥堵的目的。
Regional boundary main intersection signal control method based on deep reinforcement learning
一种基于深度强化学习的区域边界主交叉口信号控制方法
14.09.2021
Patent
Elektronische Ressource
Chinesisch
IPC: | G06F ELECTRIC DIGITAL DATA PROCESSING , Elektrische digitale Datenverarbeitung / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen / G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS |
Urban intersection signal control method based on deep reinforcement learning
Europäisches Patentamt | 2024
|Multi-intersection signal cooperative control method based on deep reinforcement learning
Europäisches Patentamt | 2025
|Single-intersection signal control method based on deep reinforcement learning algorithm
Europäisches Patentamt | 2023
|Multi-intersection traffic signal control method based on deep reinforcement learning
Europäisches Patentamt | 2023
|Multi-intersection traffic signal control method based on deep reinforcement learning
Europäisches Patentamt | 2024
|