Traffic signal control method based on Q learning

Abstract

The invention, which relates to the technical field of traffic control engineering, discloses a traffic signal control method based on Q learning. The method is characterized in that the method comprises the following steps: carrying out information collection; and for each moment, acquiring queuing vehicle information of all lanes of the intersection. According to the traffic signal control method based on Q learning, a Q value table is trained by using Q learning method, so that different Q values of different lamp states are performed under the condition that the lane queuing vehicles are given; a lamp state action which enables the Q value table to be maximum is selected and is used for signal control according to the trained Q value table; and finally, the control of the traffic signal can adapt to the randomness of the traffic network, so that better control efficiency is achieved. Problems that in the prior art, traffic flow collection of multiple sections of crossroads is difficult to achieve through the existing calculation mode, randomness of a road network is ignored by signal control of calculated time, the road network cannot adapt to the current urban road traffic situation, and consequently the vehicle passing efficiency is low and the delay time is long are solved.

Description

technical field [0001] The invention relates to the technical field of traffic control engineering, in particular to a traffic signal control method based on Q learning. Background technique [0002] The invention of automobile has brought great convenience to people's travel and life, but with the development of social economy, the rapid increase in the number of automobiles in cities has brought a series of traffic problems, among which the problem of traffic congestion is particularly prominent. Adopting a reasonable signal control scheme is an effective method to relieve traffic congestion. At present, most traffic lights in the city are controlled by timing, that is, the traffic phase and length are pre-calculated according to the historical traffic flow, so that the lights are switched at a fixed time. [0003] According to an improved traffic signal control method based on Q-learning proposed in Chinese Patent Application Publication No. CN 105654744 A, the improved ...

AI Technical Summary

Problems solved by technology

[0005] Aiming at the deficiencies of the prior art, the present invention provides a traffic signal control method based on Q-learning, which has the advantages of being able to adapt to the randomness of the traffic road network, etc., and solves the problem that the calculation method in the prior art is difficult to achieve the traffic flow at multi-section intersections Convergence, the signal control of the calculated time ignores the randomness of the road network and cannot adapt to the current urban road traffic conditions, resulting in low vehicle traffic efficiency and long delays

Images