Traffic signal control method, system and medium

Abstract

The invention provides a traffic signal control method. The method includes the following steps: a data generation step: generating expert data; a network structure construction step: constructing a behavior strategy evaluation network structure; an evaluation method construction step: constructing a behavior strategy evaluation method; a network loss function construction step: constructing a behavior strategy loss function; an acquisition step: acquiring behavior strategy information; an evaluation network loss function construction step: constructing an evaluation network loss function; a timing difference value acquisition step: calculating a timing difference value according to the evaluation network loss function; a behavior updating step: updating a behavior strategy according to the timing difference value; and a prediction result calculation step: obtaining a prediction result and applying the prediction result to the traffic. By adopting the scheme of the invention, the exploration cost of the system in the state / behavior space is effectively reduced; and the convergence speed and predictive control performance are improved.

Description

technical field [0001] The present invention relates to the field of computer software and traffic, and in particular, to a traffic signal control method, system and medium. In particular, it relates to a traffic signal control method based on expert data and behavior strategy evaluation framework. Background technique [0002] Traffic congestion has become one of the main problems affecting everyone's daily life, and people often spend a lot of time on the road. An effective way to alleviate this problem is to develop an efficient traffic light control system. [0003] With the development of computing resources and machine learning algorithms, more and more researchers have begun to use reinforcement learning (RL) algorithms to study traffic signal control methods. In traffic signal control, reinforcement learning is to learn an optimized behavior policy through the interaction between the agent and the environment, and make this policy meet the needs of realistic traffi...

AI Technical Summary

Problems solved by technology

However, in terms of traffic signal control, there is no combination of expert data and behavioral policy evaluation framework to solve the problem of traffic signal control, which is challenging and has an important impact on people's daily life.

Images