Research Article
Trajectory Optimization of CAVs in Freeway Work Zone considering Car-Following Behaviors Using Online Multiagent Reinforcement Learning
| | | Average delay (s) | Travel time (h) | Average speed (km/h) | Average stops | Stop delay (s) |
| | Scenario 0 | 63.412 | 36.234 | 40.030 | 6.661 | 9.035 | | Scenario 1 (% diff. vs. scenario 0) | 57.979 (−8.6) | 35.160 (−2.9) | 42.591 (6.4) | 5.602 (−15.9) | 6.235 (−3.1) | | Scenario 2 (% diff. vs. scenario 1) | 55.451 (−4.4) | 33.218 (−6.4) | 43.876 (3.0) | 4.994 (−10.9) | 5.404 (−13.3) | | Scenario 3 (% diff. vs. scenario 2) | 50.327 (−9.2) | 26.128 (−21.3) | 48.697 (10.9) | 3.533 (−29.3) | 3.923 (−27.4) |
|
|
