Signal control device and signal control method based on reinforcement learning

Abstract

Provided are a signal control device and a signal control method. According to one embodiment disclosed in the present specification, a signal control device for controlling a traffic signal in a crossroad on the basis of a reinforcement learning model may include an image capturing unit that captures an image of each of a plurality of crossroads to acquire a plurality of crossroad images; a storage unit that stores the programs for controlling the signals; and a control part having at least one processor for executing the program to calculate the control information for controlling a signal light of each of a plurality of crossroads by using the crossroad images acquired by the image capturing section. The control unit calculates, by using a plurality of agents based on the trained reinforcement learning model, the control information for controlling the traffic light for each of the plurality of crossroads on the basis of the motion information calculated by the plurality of agents by using state information calculated on the basis of each of the plurality of crossroad images as an input, wherein the reinforcement learning model is trained to output the motion information for controlling a signal lamp by using the state information and rewards as input values.

Description

technical field [0001] The embodiments disclosed in this specification relate to a signal control device and a signal control method based on reinforcement learning, and in more detail, relate to a device and a method for controlling traffic signals in a plurality of intersections. Background technique [0002] Recently, as the number of people buying vehicles for convenience or work reasons has increased, so has the number of vehicles on the road. Due to the increase of these vehicles, traffic jams are also increasing, and traffic jams may occur due to various factors such as road environments, driver conditions, vehicle breakdowns, and vehicle accidents. [0003] One of the causes of traffic jams is the problem of the traffic signal system in the road environment. For example, traffic signals control the flow of vehicles, and since they determine the direction of traffic of vehicles at regular intervals, when the number of vehicles in a particular direction increases, tra...

AI Technical Summary

Problems solved by technology

However, in the above-mentioned prior art, the artificial intelligence model is only used as a means to detect whether there is a vehicle on a specific lane through simple image analysis, and the determination of the next signal based on the detected information is through the existing fragmented calculation. Therefore, there is a problem that it is difficult to improve the efficiency of the signal system

Images