论文信息 - FREEWAY TRAFFIC CONTROL USING Q-LEARNING

FREEWAY TRAFFIC CONTROL USING Q-LEARNING

In this paper, the standard Q-learning algorithm is applied to control the density of an arbitrary freeway via ramp metering in a macroscopic level. The reinforcement learning algorithms have proven to be effective tools for letting an agent learn from its experiences generated by its interaction with an environment. The performance of the algorithm as well as its robustness against communication failure is studied. The results of the simulations demonstrated the effectiveness the technique.

Andreas Hegyi | Mohsen Davarynejad | Jos Vrancken | Yubin Wang

[1] B. De Moor,et al. Anticipative model predictive control for ramp metering in freeway networks , 2003, Proceedings of the 2003 American Control Conference, 2003..

[2] Bart De Schutter,et al. Model predictive control for optimal coordination of ramp metering and variable speed limits , 2005 .

[3] Andras Hegyi,et al. Model predictive control for integrating traffic control measures , 2004 .

[4] Louis Wehenkel,et al. Reinforcement Learning Versus Model Predictive Control: A Comparison on a Power System Problem , 2009, IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics).

[5] Richard S. Sutton,et al. Learning to predict by the methods of temporal differences , 1988, Machine Learning.

[6] Andrew W. Moore,et al. Reinforcement Learning: A Survey , 1996, J. Artif. Intell. Res..

[7] Richard S. Sutton,et al. Introduction to Reinforcement Learning , 1998 .

[8] Jan M. Maciejowski,et al. Predictive control : with constraints , 2002 .

[9] Markos Papageorgiou,et al. METANET: A MACROSCOPIC SIMULATION PROGRAM FOR MOTORWAY NETWORKS , 1990 .