In this thesis, a comprehensive high-level autonomous driving control pipeline, along with an optimization-based motion planning algorithm, is proposed, detailed, and implemented into BCNeMotorsport’s Autonomous Systems Control software stack. The main goal of the presented algorithms is to improve the competitiveness of BCN eMotorsport’s autonomous vehicle, the CAT15X, reaching levels of performance and reliability never seen throughout the team’s history. In section 3 a Nonlinear Model Predictive Controller is proposed in order to handle the lateral behaviour of the vehicle using the curvature-based dynamic bicycle model presented in section 2.3. In section 4 an Adaptive Cruise Controller is proposed in order to handle the longitudinal behaviour of the car, closely following the self-computed velocity profile defined in section 4.2. Finally, an offline Optimization-based Motion Planning algorithm is proposed in section 5. This algorithm utilizes the dynamic bicycle model, introduced in section 2.3, as its predictive basis. Theimplementationprocedureaswellasthesimulationresultsforallthepresentedalgorithms are also specified. In addition, the real testing performance of the proposed control pipeline is detailed in section 6. The complete software designed and implemented within this thesis can be found in: • Lateral MPC • Cruise Controller • Trajectory Optimizer
Motion planning and control pipeline for a Formula Student autonomous vehicle.
2024-01-31
ETSEIB-240.184149
Theses
Electronic Resource
English
DDC: | 629 |
MPC Controller for Autonomous Formula Student Vehicle
British Library Conference Proceedings | 2020
|MPC Controller for Autonomous Formula Student Vehicle
British Library Conference Proceedings | 2020
|MPC Controller for Autonomous Formula Student Vehicle
SAE Technical Papers | 2020
|Path Following Control for Autonomous Formula Racecar: Autonomous Formula Student Competition
British Library Conference Proceedings | 2017
|