แจ้งเอกสารไม่ครบถ้วน, ไม่ตรงกับชื่อเรื่อง หรือมีข้อผิดพลาดเกี่ยวกับเอกสาร ติดต่อที่นี่ ==>
หากไม่มีอีเมลผู้รับให้กรอก thailis-noc@uni.net.th
หากไม่มีอีเมลผู้รับให้กรอก thailis-noc@uni.net.th
Khin Khin Kyi. Low spee car parking manoeuvring using deep reinforcement learning. Master's Degree(Automotive and Advanced Transportation Engineering). King Mongkut's Institute of Technology Ladkrabang. KMITL Lifelong Learning Center. : King Mongkut's Institute of Technology Ladkrabang, 2024.
Low spee car parking manoeuvring using deep reinforcement learning
Name: Khin Khin Kyi
LCSH:
Automated vehicles
-- Control
LCSH:
Motor vehicle driving
Abstract:
Due to the advancement of control systems, numerous research on autonomous driving has been conducted in different driving scenarios and environments intended to change to fully autonomous mode driving systems. Research on autonomous driving has been explored using various driving algorithms in diverse driving conditions. In recent years, Machine Learning and Deep Learning algorithms have been employed in automotive research field as a result of their astonishing performance in different industries. This study focuses on the parking scenario for a four-wheel vehicle in the Carla 3D Simulation Environment, using Deep Reinforcement Learning. Carla simulator is a good 3D simulator workspace as it provides realistic driving experience including a variety of infrastructures and non-player characters. The agent vehicle undergoes simultaneous learning and training in the environment, rewards and penalties are used as feedback results to evaluate the behavior of each action in every episode. The inputs to the DQN network are the state conditions of the agent in the environment and DQN network model generates the most suitable action for the given state based on the trained experiences. The collision sensor is used to detect collision accidents and three test scenarios are conducted with different spawn conditions and parking destination to evaluate the agent performance in each scenario. The agent successfully accomplished the parking process in different testing conditions and earned high rewards in each training episode by parking the car at the designated location without any collisions in a short amount of time
King Mongkut's Institute of Technology Ladkrabang. KMITL Lifelong Learning Center
Address:
BANGKOK
Email:
lifelong@kmitl.ac.th
Name: Supat Kittiratsatcha
Role:
Thesis Advisor
Email :
supat.ki@kmitl.ac.th
Name: Masaki Yamakita
Role:
Thesis Co-Advisor
Created:
2024
Modified:
2025-10-02
Issued:
2025-10-02
วิทยานิพนธ์/Thesis
application/pdf
eng
DegreeName: Master of Engineering
Level: Master's Degree
©copyrights King Mongkut's Institute of Technology Ladkrabang
RightsAccess:
ใช้เวลา
0.039545 วินาที
Khin Khin Kyi
| Title | Contributor | Type |
|---|---|---|
| Low spee car parking manoeuvring using deep reinforcement learning
สถาบันเทคโนโลยีพระจอมเกล้าเจ้าคุณทหารลาดกระบัง Khin Khin Kyi | Supat Kittiratsatcha Masaki Yamakita | วิทยานิพนธ์/Thesis |
Supat Kittiratsatcha
| Title | Creator | Type and Date Create |
|---|---|---|
| Development of anti-islanding hybrid detection technique in grid-connected distributed generation
มหาวิทยาลัยเชียงใหม่ Supat kittiratsatcha;Suttichai Premrudeepreechacharn;Yuttana Kumsuwan;Peerapol Jirapong;Kasemsak Uthaichana | Manop Yingram | วิทยานิพนธ์/Thesis |
| Influence on partial-flow particulate on diesel vehicle thermal efficiency and emission characteristics using B20
สถาบันเทคโนโลยีพระจอมเกล้าเจ้าคุณทหารลาดกระบัง Supat Kittiratsatcha;Kobsak Sriprapha;Katsunori Hanamura | Plan Teekatasn Cosh | วิทยานิพนธ์/Thesis |
| Low spee car parking manoeuvring using deep reinforcement learning
สถาบันเทคโนโลยีพระจอมเกล้าเจ้าคุณทหารลาดกระบัง Supat Kittiratsatcha;Masaki Yamakita | Khin Khin Kyi | วิทยานิพนธ์/Thesis |
Masaki Yamakita
| Title | Creator | Type and Date Create |
|---|---|---|
| Low spee car parking manoeuvring using deep reinforcement learning
สถาบันเทคโนโลยีพระจอมเกล้าเจ้าคุณทหารลาดกระบัง Supat Kittiratsatcha;Masaki Yamakita | Khin Khin Kyi | วิทยานิพนธ์/Thesis |