| dc.contributor | 資訊系 | |
| dc.creator (作者) | 張宏慶 | |
| dc.creator (作者) | Jang, Hung-Chin;Lee, Chun-Yee | |
| dc.date (日期) | 2022-12 | |
| dc.date.accessioned | 16-Feb-2024 15:36:37 (UTC+8) | - |
| dc.date.available | 16-Feb-2024 15:36:37 (UTC+8) | - |
| dc.date.issued (上傳時間) | 16-Feb-2024 15:36:37 (UTC+8) | - |
| dc.identifier.uri (URI) | https://nccur.lib.nccu.edu.tw/handle/140.119/149878 | - |
| dc.description.abstract (摘要) | In recent years, most research on urban roadside parking space search has focused on improving the prediction of the vacancy of roadside parking spaces. One simple but expensive practice is setting up sensors in each parking space to provide drivers with realtime parking space information so drivers can find suitable parking spaces. Although providing realtime information on each parking space can help drivers when choosing a driving route, there is a possibility that other drivers take the parking space during the time of getting to the specific parking space. A better approach to the parking space search is to find a suitable parking area rather than specific parking spaces. Predicting the probability of an available parking area can reduce the time the vehicle lingers in search of a parking space. In this study, we proposed to use Deep Q-Learning with fewer sensors to solve the problem. Besides, we used LSTM (Long Short-Term Memory) and GRU (Gated Recurrent Unit) models to improve the accuracy in estimating the Q value of the Deep Q-learning. Finally, we compared the performance of Q-Learning and Deep Q-Learning using simulated traffic flow data. | |
| dc.format.extent | 774689 bytes | - |
| dc.format.mimetype | application/pdf | - |
| dc.relation (關聯) | 2022 International Conference on Computational Science and Computational Intelligence (CSCI), American Council on Science and Education | |
| dc.title (題名) | Study on Q-Learning and Deep Q-Learning in Urban Roadside Parking Space Search | |
| dc.type (資料類型) | conference | |
