This paper provides a data-driven model-based solution for control problem of path following for autonomous vehicles. The modeling of the system is based on the Linear Parameter-Varying (LPV) framework, but the selections of the scheduling variables and the LPV model parameters are based on machine learning methods. The advantage of the method is that the performances of the system can be guaranteed, while the generated model is valid in an extended operation range. The control design is based on the LPV method, in which the novel vehicle model is incorporated. The effectiveness of the method is illustrated through comparative simulation scenarios in CarMaker simulation environment.
Paper
Full text
LPV based data-driven modeling and control design for autonomous vehicles
Semantic Scholar · Engineering · 2020
Abstract
This paper provides a data-driven model-based solution for control problem of path following for autonomous vehicles. The modeling of the system is based on the Linear Parameter-Varying (LPV) framework, but the selections of the scheduling variables and the LPV model parameters are based on machine learning methods. The advantage of the method is that the performances of the system can be guaranteed, while the generated model is valid in an extended operation range. The control design is based on the LPV method, in which the novel vehicle model is incorporated. The effectiveness of the method is illustrated through comparative simulation scenarios in CarMaker simulation environment.