This paper presents a planning and control pipeline for an autonomous race car to drive around a track that may not be previously known for three laps. In the case of a limited perception range in the first lap, boundary detection and motion planning separately extract and optimize the trajectory to minimize the trajectory curvature. After finishing the mapping, multi-strategy NMPC is used to optimize or track the trajectory. We use the real-world map data from Formula Student Autonomous China 2019 for experiments. The experiment shows that under the same vehicle model, the system can significantly improve the performance of the race car.
Real-Time Motion Planning and Control for a Formula Student Driverless Car
Lect. Notes Electrical Eng.
2022-01-13
17 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Real-Time Motion Planning and Control for a Formula Student Driverless Car
British Library Conference Proceedings | 2022
|Startklar fuer die Formula Student Driverless
Kraftfahrwesen | 2016
|AUTONOMOUS DRIVING SYSTEM DESIGN FOR FORMULA STUDENT DRIVERLESS RACECAR
British Library Conference Proceedings | 2018
|