Intelligent traffic light system based on deep learning and method for controlling traffic light

Abstract

The invention discloses an intelligent traffic light system based on deep learning and a method for controlling a traffic light. The intelligent traffic light system comprises a data collecting and processing module, a deep learning module and a traffic light central controlling module, wherein the data collecting and processing module is used for collecting images of an intersection corresponding to the traffic light in real time, and identifying vehicle model information, a traffic volume, the vehicle queue length and the corresponding passing time of vehicles; the deep learning module comprises a neural network which can simulate a human brain for analysis and learning, and a control model, and the control model inputs the vehicle model information, the traffic volume, the vehicle queue length and the corresponding passing time of the vehicles into the neural network for autonomous deep learning and forming the control model; the traffic light central controlling module is used for controlling the traffic light. According to the intelligent traffic light system based on deep learning, the control model which is formed through autonomous deep learning can calculate and control the duration of a green light after the next change of the traffic light according to the vehicle queue length at the intersection and the vehicle model information, and efficiently solve the problem of heavy traffic.

Description

technical field [0001] The invention relates to the field of traffic lights, in particular to an intelligent traffic light system based on deep learning and a method for controlling traffic lights. Background technique [0002] Current traffic lights adopt PLC control program to control the time of traffic lights, that is, the PLC control program of traffic lights is set in advance according to the situation of the crossing. For example, in the period of a crossroads, according to the north-south direction, the red light is on for 25 seconds, turn to the green light for 25 seconds, and then flash 3 times according to the rule of once every second, then turn the yellow light on for 2 seconds, and the east-west direction green light is on for 20 seconds seconds, then flash 3 times, turn to yellow light for 2 seconds, then red light for 30 seconds, complete a cycle, and so on. [0003] The passing time of green lights in this fixed cycle and fixed direction will face a problem...

AI Technical Summary

Problems solved by technology

[0003] The passing time of green lights in this fixed cycle and fixed direction will face a problem: it is impossible to change the time of traffic lights according to the traffic flow. It often occurs in the peak congestion stage. Some roads have little traffic but the traffic lights take too long. Or the traffic light time is too short when there is too much traffic

In the prior art, the regulation of traffic lights through PLC fixed programs will affect the actual traffic efficiency, and the actual traffic conditions on the road surface cannot be well regulated.

Images