The urban traffic problem has become more and more serious with the development of social and economic development. Researches are being carried out in many directions to solve this problem. In particular, smart traffic control systems that can effectively manage urban traffic without increasing infrastructure are being widely studied.
However, due to the characteristics peculiar to the traffic system itself, there arise many difficulties in building an intelligent traffic system. And it is difficult to realize due to the fact that the amount of data related to its decision-making are vast and artificial. In order to overcome this, artificial intelligence techniques such as multi-agent systems (MAS), deep learning, and Q learning are being actively applied.
With the advent of AI technology, distributed autonomous control was made possible, and MAS technology has become an important technology to solve the traffic coordination problem. From the viewpoint of area traffic coordination control, it is particularly important to adopt more intelligent control strategies at urban intersections where most traffic congestions usually occur.
Kim Thae Yong, a researcher at the Faculty of Automatics, proposed a Multi-Agent architecture by which agents (intersections) can implement traffic signal control and coordination control over intersections, mutually utilizing the characteristics of the low-dimensioned traffic state of other agents within the traffic networks, thus improving the performance of area traffic coordination.
He simulated the algorithm in several possible cases and the control strategy for coordination control by means of VISSIM. The simulation results show that the proposed method can reduce the total traffic time and queue length under a congested traffic environment.
For more details, you can refer to his paper “Coordination control of area traffic networks with multi-agent architecture based on deep recurrent Q-learning networks” in “Second International Conference on Intelligent Transportation and Smart Cities (ICITSC 2025)” (EI).