Exploring the potential of Dyna-Q learning for multi-agent systems to solve multi-intersection traffic network problems
Relieving urban traffic congestion has always been an urgent caU in a dynamic traffic network. This research aims to control the traffic flow within a traffic network consists of two signalized intersections with traffic ramp. The massive traffic network problem is dealt through dynamic Q-Iearning (...
Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Research Report |
| Language: | English |
| Published: |
Universiti Malaysia Sabah
2012
|
| Subjects: | |
| Online Access: | https://eprints.ums.edu.my/id/eprint/22528/1/Exploring%20the%20potential%20of%20Dyna-Q%20learning%20for%20multi-agent%20systems%20to%20solve%20multi-intersection%20traffic%20network%20problems.pdf https://eprints.ums.edu.my/id/eprint/22528/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Universiti Malaysia Sabah |
| Language: | English |
| Summary: | Relieving urban traffic congestion has always been an urgent caU in a dynamic traffic network. This research aims to control the traffic flow within a traffic network consists of two signalized intersections with traffic ramp. The massive traffic network problem is dealt through dynamic Q-Iearning (Dyna-Q) actuated traffic signalisation. where the traffic phases will be monitored as immediate actions can be accomplished during congestion to minimise the number of vehicles in queue. The simulation results show the total vehicles passed through the network with proposed algorithm are 2.9 - 19.0 % more than the existing pre-timed traffic signalisation due to its flexibility in changing the traffic signal timing plan according to the traffic conditions and necessity. |
|---|