基于Agent控制器的单路口信号灯学习控制方法研究

Agent-based Learning Control Method for Single Intersection Traffic Lights

将强化学习中的Q学习算法与Agent技术相结合,构成Agent控制器对单路口的信号灯进行控制。介绍了将经验知识与Q学习算法相结合实现的Agent学习机制,提出了一种适合交通环境的强化函数,以解决单路口的动态实时控制。经仿真软件Paramics对实际路口进行仿真实验,验证了该方法具有较好的控制效果。

In this paper, Agent technology was combined with Q-learning to make an Agent controller, which controlled the traffic lights in single intersection. The Agent learning mechanism was realized by combining experience knowledge with Q-learning. A reward function for real traffic condition was proposed to control traffic dynamically and on real-time. The performance of this method is evaluated by the traffic simulation software Paramics, and the experimental results indicate the effectiveness of the method.