It is important for autonomous vehicles to plan at the unsignalized intersection due to the uncertainty of visible vehicles and phantom vehicles in the blind zone. This paper proposes a hierarchical framework to optimize the left-turn planning at the intersection to solve this problem. The framework is composed of a higher-level candidate paths selector and a lower-level Partially Observable Markov Decision Process planner. All possible vehicle distributions in the blind zone are represented in the form of a set, which makes the method have the good universality. The simulation experiments show that autonomous vehicle can pass the intersection safely and efficiently.
A hierarchical planning framework of the intersection with blind zone and uncertainty
2021-09-19
522890 byte
Conference paper
Electronic Resource
English
Unmanned vehicle blind area intersection planning method considering uncertainty
European Patent Office | 2023
|INTERSECTION BLIND-GUIDING SYSTEM, BLIND-GUIDING METHOD AND BLIND-GUIDING TERMINAL
European Patent Office | 2020
|Trajectory Planning Framework Combining Replanning and Hierarchical Planning
Transportation Research Record | 2023
|