TRAFFIC LIGHT CONTROL SYSTEM SIMULATOR FOR REINFORCEMENT LEARNING HARDWARE ACCELERATOR
Traffic light regulation is very crucial in urban areas. The more efficient the regulatory system, the more benefits will be obtained. Reinforcement Learning is an algorithm that will be tested to become the basis for the traffic light control system. Testing is done by modeling the intersection...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/66471 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| id |
id-itb.:66471 |
|---|---|
| spelling |
id-itb.:664712022-06-28T12:22:09ZTRAFFIC LIGHT CONTROL SYSTEM SIMULATOR FOR REINFORCEMENT LEARNING HARDWARE ACCELERATOR Haritsa Maulana, Daffa Indonesia Final Project Hardware Accelerator,Traffic Light control system,,Reinforcement learning, simulator. INSTITUT TEKNOLOGI BANDUNG https://digilib.itb.ac.id/gdl/view/66471 Traffic light regulation is very crucial in urban areas. The more efficient the regulatory system, the more benefits will be obtained. Reinforcement Learning is an algorithm that will be tested to become the basis for the traffic light control system. Testing is done by modeling the intersection scheme with certain parameters using the python language and software. The simulator is getting action from hardware accelerator and give the action to the environment. The simulator also inteperate the congestion condition and convert it to state. Comparison between Reinforcement Learning method and fixed-time method was also conducted. From the results of this final project, it is found that using the reinforcement learning method can reduce congestion by up to 15% when compared to the fixed-time method. text |
| institution |
Institut Teknologi Bandung |
| building |
Institut Teknologi Bandung Library |
| continent |
Asia |
| country |
Indonesia Indonesia |
| content_provider |
Institut Teknologi Bandung |
| collection |
Digital ITB |
| language |
Indonesia |
| description |
Traffic light regulation is very crucial in urban areas. The more efficient the
regulatory system, the more benefits will be obtained. Reinforcement Learning is
an algorithm that will be tested to become the basis for the traffic light control
system. Testing is done by modeling the intersection scheme with certain
parameters using the python language and software. The simulator is getting action
from hardware accelerator and give the action to the environment. The simulator
also inteperate the congestion condition and convert it to state. Comparison
between Reinforcement Learning method and fixed-time method was also
conducted. From the results of this final project, it is found that using the
reinforcement learning method can reduce congestion by up to 15% when
compared to the fixed-time method. |
| format |
Final Project |
| author |
Haritsa Maulana, Daffa |
| spellingShingle |
Haritsa Maulana, Daffa TRAFFIC LIGHT CONTROL SYSTEM SIMULATOR FOR REINFORCEMENT LEARNING HARDWARE ACCELERATOR |
| author_facet |
Haritsa Maulana, Daffa |
| author_sort |
Haritsa Maulana, Daffa |
| title |
TRAFFIC LIGHT CONTROL SYSTEM SIMULATOR FOR REINFORCEMENT LEARNING HARDWARE ACCELERATOR |
| title_short |
TRAFFIC LIGHT CONTROL SYSTEM SIMULATOR FOR REINFORCEMENT LEARNING HARDWARE ACCELERATOR |
| title_full |
TRAFFIC LIGHT CONTROL SYSTEM SIMULATOR FOR REINFORCEMENT LEARNING HARDWARE ACCELERATOR |
| title_fullStr |
TRAFFIC LIGHT CONTROL SYSTEM SIMULATOR FOR REINFORCEMENT LEARNING HARDWARE ACCELERATOR |
| title_full_unstemmed |
TRAFFIC LIGHT CONTROL SYSTEM SIMULATOR FOR REINFORCEMENT LEARNING HARDWARE ACCELERATOR |
| title_sort |
traffic light control system simulator for reinforcement learning hardware accelerator |
| url |
https://digilib.itb.ac.id/gdl/view/66471 |
| _version_ |
1822933048920899584 |