SARSA-based trunk line traffic control optimization method

Abstract

The invention discloses an SARSA learning-based trunk traffic control optimization method. The method comprises the following steps: 1) for a period of continuous time, acquiring traffic light state information of a target trunk intersection and queuing vehicle information of a lane corresponding to each traffic light; 2) integrating the initial trunk line-lamp state data to obtain a trunk line-lamp state data set; 3) optimizing a Q value table until convergence according to the trunk line-lamp state data set collected in the step 1); and 4) after the Q value table converges to satisfactory precision, obtaining a trunk line traffic optimization scheme based on SARSA learning. Compared with the prior art, the method has the advantages that hidden information contained in the vehicles in thetrunk line is mined through an artificial intelligence method, and the obtained trunk line timing scheme has higher traffic efficiency than a traditional green wave band control algorithm under the condition of a traffic flow road network with high randomness.

Description

technical field [0001] The invention relates to the fields of traffic control engineering and artificial intelligence application, in particular to a SARSA (State, Action, Reward, Next State, Next Action) learning method and a trunk line traffic control optimization method. Background technique [0002] In the urban road network, the optimal control of arterial traffic is often the focus of determining the residents' traffic experience. But in recent years, with the popularization of automobiles, road network facilities cannot keep up with the rapid development of traffic flow. Previously designed signal control methods for trunk lines such as green wave band and two-way green wave band can no longer adapt to the current traffic flow of the trunk line. The main reason is that the above methods are all given timing schemes and cannot handle traffic Random changes in traffic. The SARSA-based arterial traffic control optimization method designed in this patent is a real-time ...

AI Technical Summary

Problems solved by technology

[0003] In order to solve the problem that the arterial traffic timing scheme cannot well adapt to random changes in traffic flow, based on the real-time acquired arterial traffic data, the present invention proposes a traffic signal timing optimization method using vehicle queuing data at each intersection of the arterial line. The traffic signal control optimization scheme based on SARSA proposed by traffic can improve the efficiency of traffic flow, and can deal with the random change of complex traffic flow

Images