The invention discloses an intersection traffic signal control method based on fuzzy reinforcement learning. The method comprises the following steps: constructing a reinforcement learning signal lamp intelligent agent by setting a vehicle state variable, a signal lamp action variable and a reward function; inputting state data of vehicles arriving at the intersection, and executing a fuzzy reasoning strategy to select an output signal lamp action; obtaining new vehicle state data and a current reward value, and putting the data into an experience playback pool; when the data of the experience playback pool reaches a certain capacity, extracting a batch of data for training until a stop condition is reached; and using the trained reinforcement learning signal lamp intelligent agent to perform traffic signal control of the intersection. According to the method, the defect that reinforcement learning model training is difficult to converge and unstable is overcome, and the problem that vehicle waiting time at the intersection is too long is solved.
本发明公开一种基于模糊强化学习的交叉口交通信号控制方法,所述方法包括通过设置车辆状态变量、信号灯动作变量以及奖励函数来构建强化学习信号灯智能体;输入到达交叉口的车辆状态数据,并执行模糊推理策略选择输出信号灯动作;得到新的所述车辆状态数据和当前奖励值,并将数据放入经验回放池;待所述经验回放池的数据达到一定容量,抽取一批数据进行训练直至达到停止条件;使用训练好的所述强化学习信号灯智能体进行所述交叉口的交通信号控制。该方法克服了强化学习模型训练难以收敛不稳定的弊端,解决了交叉口车辆等待时间过长的问题。
Intersection traffic signal control method based on fuzzy reinforcement learning
一种基于模糊强化学习的交叉口交通信号控制方法
2024-09-27
Patent
Electronic Resource
Chinese
Multi-intersection traffic signal control method based on deep reinforcement learning
European Patent Office | 2023
|Multi-intersection traffic signal control method based on deep reinforcement learning
European Patent Office | 2024
|European Patent Office | 2025
|European Patent Office | 2024
|