Method, system and device for controlling single-intersection traffic signal control based on deep reinforcement learning

Abstract

The invention belongs to the field of urban traffic control, and particularly relates to a method, a system and a device for controlling single-intersection traffic signal control based on deep reinforcement learning, which is to solve the problem that the traffic signal control effect of complex traffic conditions is not good. The method comprises the steps of: establishing a microscopic trafficsimulation environment, defining parameters, and setting a determination network and a traffic signal generation network; calculating the training error of the network based on the data of the currentstage and the previous stage, and updating the network parameters; based on the updated determination network and the data of the current stage and the previous stage, calculating the training errorof the updated determination network, and updating the parameters of the determination network and the traffic signal generation network; and adopting the trained traffic signal generation network toobtain the next phase length of a signal lamp at the intersection. The method, the system and the device for controlling single-intersection traffic signal control based on deep reinforcement learningreduces the research work of knowing the traffic information of the intersection beforehand, and can make adjustments in time as the traffic demand of the intersection is changed, thereby greatly improving the effect of traffic signal control of complex traffic conditions.

Description

technical field

[0001] The invention belongs to the field of urban traffic control, and in particular relates to a single intersection traffic signal control method, system and device based on deep reinforcement learning. Background technique

[0002] Traffic signal control is an important means of urban traffic management and control. A reasonable traffic signal control strategy can not only improve the operational efficiency of the traffic system, but also effectively reduce the occurrence of traffic accidents. The short-term traffic demand at road intersections has the characteristics of time-varying, nonlinear, and complexity, and it is difficult to establish an accurate mathematical model. Simple timing control and induction control methods are difficult to adapt to the dynamic, complex, and rapid changes in traffic flow. Ineffective.

[0003] The deep reinforcement learning method integrates deep learning and reinforcement learning technology, combines the feature re...

Images