A Regional Traffic Signal Control Strategy with Deep Reinforcement Learning

In this paper, a deep reinforcement learning algorithm for regional intersection traffic control is proposed for improves the capacity of the road network. This method interacts with the traffic environment to learn the optimal signal control strategy by reinforcement learning. And it using deep learning to reduce the dimension of the huge traffic flow data generated by the road network. The denoising stacked autoencoders are adopted to handle the abnormal dynamic traffic data. Simulations in platform consisting of VISSIM and Python are applied to test the algorithm. The performance of the proposed method is comprehensively compared with a traditional algorithms and a fixed signal timing method with green time difference for regional traffic signal control under different traffic demand. Simulation results suggest that the proposed method significantly reduces the average delay in the traffic network when traffic conditions changed rapidly.