The length of the planning horizon when using nonlinear model predictive control is limited by computational resources. To achieve long-horizon vehicle planning and control, the practical feasibility of applying real-time iterations is demonstrated with an experiment on an automated vehicle. The proposed implementation is real-time capable with a planning horizon of over 10s, requiring a fraction of the computational time compared to a solver for general nonlinear programming problems. The solutions obtained are close to the local optimum, even in the presence of real-world disturbances. Important details for the successful implementation in this domain are discussed and illustrated through simulation.
Long-Horizon Vehicle Planning and Control Through Real-Time Iterations
2023-06-04
1229074 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Real-time moving horizon planning and control of aerial systems under uncertainties
BASE | 2020
|SAE Technical Papers | 2019
|British Library Conference Proceedings | 2019
|Real-Time Receding-Horizon Control Algorithm for Nonlinear Systems
British Library Online Contents | 1997
|