Traffic Flow Optimization on Melbourne Parking Probability 0.1 Real-world data
99.02Avg Time Loss (s)D-PA (Ours)
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| D-PA (Ours)Parking Probability=0.12025.12 | 99.02 | 13.03 | 1.44 | |
| D-PA (DQN)Parking Probability=0.1, Backbone=DQN2025.12 | 105.36 | 7.46 | 0.85 | |
| S-PAParking Probability=0.12025.12 | 106.84 | 6.16 | 0.8 | |
| D-PA (A2C)Parking Probability=0.1, Backbone=A2C2025.12 | 106.92 | 6.09 | 0.89 | |
| D-PA (Du-DQN)Parking Probability=0.1, Backbone=Dueling DQN2025.12 | 108.69 | 4.54 | 0.71 | |
| D-PA (PPO)Parking Probability=0.1, Backbone=PPO2025.12 | 108.8 | 4.44 | 0.72 | |
| D-PA (D-DQN)Parking Probability=0.1, Backbone=Double DQN2025.12 | 109.43 | 3.89 | 0.71 | |
| C-PAParking Probability=0.12025.12 | 109.65 | 3.69 | 59.75 | |
| No-PAParking Probability=0.12025.12 | 113.85 | — | 0.48 |