Seminar 1: BrIAS Fellow Prof. Ilya Kolmanovsky
Exploiting Supervisory Schemes and the Interplay Between Computations and Closed-Loop Properties in Model Predictive Control
Abstract: Model Predictive Control (MPC) leads to algorithmically defined nonlinear feedback laws for systems with pointwise-in-time state and control constraints. These feedback laws are defined by solutions to appropriately posed optimal control/trajectory optimization problems that are (typically) solved online. There is a growing interest in the use of MPC for practical applications, including as an enabling technology for control and trajectory generation in autonomous vehicles, including in aerospace, automotive and robotics domains. To enable MPC implementation, the solutions to MPC optimization problems must be computed reliably and within the available time. After describing several motivating applications in aerospace and automotive domains, the talk will reflect on recent research by the presenter and his students/collaborators into strategies for computing solutions in optimization problems arising in receding horizon and shrinking horizon MPC formulations. These strategies include methods for solving MPC problems inexactly, and the use of add-on supervisory schemes for MPC which reduce the computational time and enlarge the constrained closed-loop region of attraction. In particular, a Computational Governor (CG) will be described which maintains feasibility and bounds the suboptimality of the MPC warm-start by altering the reference command provided to the inexactly solved MPC problem. As it also turns out, the analysis of time distributed implementation of MPC based on fixed number of optimization algorithm iterations per time step and warm-starting benefits from the application of control-theoretic tools such as the small gain theorem; intriguingly, similar tools can be exploited in “control-aware” multi-disciplinary design optimization.
Seminar 2: BrIAS Fellow Prof. Paolo Falcone
Cautious-by-design motion planning. The role of prediction
Abstract: Safety of passengers and surrounding road users is the most important challenge in the design and deployment of autonomous driving technologies. In fact, the highest Automotive Safety Integrity Level (ASIL-D) will likely be required for autonomous driving functionalities. While fulfilling such safety requirements involves special design efforts at all levels of the autonomous driving stack, in this talk we will focus on the control design of a safe motion planner in urban environments.
We will start by illustrating a model-based control design approach to vehicle motion
planning problem, in the presence of human road users (pedestrians, cyclists, human-driven vehicles). We will show that, under mild assumptions, vehicle behavior can be made cautious
in the presence of road users and guaranteed to be persistently safe. Experimental results obtained with a passenger vehicle negotiating an intersection with a simulated pedestrian, will
be shown. An important ingredient of the proposed motion planning framework is a prediction model of the surrounding traffic. In the second part of the seminar, we will illustrate our ongoing research on humans' intent prediction in traffic environments. We will show how the evolution of a traffic scene can be predicted using very simple models and motion data (position, velocity) of road users observed in similar traffic scenes.
