Nonlinear model predictive control for automotive applications
This research line includes two activities:
- Development of trajectory planning and control algorithms for autonomous vehicles, based on the nonlinear model predictive control (NMPC) approach. The NMPC algorithms allow accomplishment of different maneuvers/operations, like way-point tracking, lane center tracking, obstacle avoidance (for fixed and moving obstacles) and constraint satisfaction (e.g., road boundaries, speed limits, etc.). The NMPC algorithms are being implemented on electronic boards. Hardware-in-the-loop tests are being carried out to validate the algorithms, where the vehicle is simulated on a PC, using a detailed vehicle model, while the NMPC algorithm is implemented on the board. This activity includes the development of a decision-making module, whose task is to provide the NMPC algorithm with one or more of the following indications: maneuver to accomplish; lane-center to track; speed reference; way-points to track; prediction of the motion of other agents in the road scenario.
- Optimization and control of the powertrain line in electric vehicles (EVs). The powertrain line of an EV is a quite complex system, composed of different interconnected elements like the battery pack, power electronics, electric motor, sensors and control systems. The goal of this research is to develop effective management algorithms for such a system, allowing the vehicle to save significant amounts of energy and reduce the emitted pollution, without significantly affecting the vehicle performance in terms of speed acceleration and comfort.
- PE7_1 Control engineering,
- PE7_3 Simulation engineering and modelling
- PE1_19 Control theory and optimisation
- Predictive control
- Automotive electronics Electric vehicles
- Trajectory planning/control