Abstract
Connected autonomous vehicles can use vehicle-to-everything (V2X) communication for energy-efficient driving. However, in urban mixed traffic involving both autonomous and human-driven vehicles, driver uncertainty and the spatio-temporal coupling of consecutive traffic signals significantly increase the complexity of eco-driving control. To address these challenges, this study proposes a hierarchical green-wave speed-based predictive eco-driving strategy for connected autonomous vehicles at successive signalized intersections, enabled by cloud-to-vehicle coordination. The upper cloud layer performs green-wave speed planning in the distance domain using direct projection algorithms to handle spatio-temporal constraints across multiple intersections. The lower vehicle layer integrates an Informer network-based speed prediction model to mitigate disturbances from surrounding traffic and refine the planned trajectory in real time. Furthermore, a speed-tracking controller based on the Pontryagin maximum principle and a binary search method is designed to achieve smooth, energy-efficient tracking performance. Simulation experiments on a Speedgoat platform demonstrate the feasibility and superior energy-saving performance of the proposed control strategy under mixed-traffic conditions.
| Original language | English |
|---|---|
| Article number | 149310 |
| Journal | Journal of Cleaner Production |
| Volume | 576 |
| DOIs | |
| Publication status | Published - 20 Sept 2026 |
| Externally published | Yes |
Keywords
- Connected autonomous vehicle
- Mixed traffic flow
- Model predictive control
- Speed planning and tracking
Fingerprint
Dive into the research topics of 'A predictive cruise eco-driving control strategy for connected autonomous vehicles through successive signalized intersections in mixed urban traffic flow'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver