The doctoral research groups ACDC and DECISIO

The DECISIO team studies methods and tools for the automatic decisions taking, in interaction with the human operator, for aerospace systems and, more generally, robotic systems. Research focuses on:

• methods for research, optimization and action planning, and vehicles or sets of vehicles missions supervision in dynamic and uncertain environments;

• the generic character of control architectures and the guarantee of safety and security of functions and architectures;

• the intrinsic performance of operators (pilots, ground station operators) in interaction with each other and with such systems, in particular through the study and modeling of cognitive functioning (learning processes, deleterious mental states) in the performance of driving or piloting tasks;

• systems engineering methods with a view to understanding the integration and interaction of different approaches in the design and production phases, as well as in operation and maintenance.

The theoretical tools used concern knowledge modeling, optimization, sequential decision-making in the presence of uncertainty, and human-machine interaction; they are based on model-based methods (logic, constraints, probabilities, etc.), machine learning methods (regression, statistics, etc.), and cognitive science.

Research activities also include experiments on simulators, robotic land and air vehicles, and aircraft.

Applications concern vehicles and groups of vehicles, primarily in the aeronautics and space sectors (aircraft and air transportation systems, helicopters, drones, orbital systems) but also terrestrial robots.

The ACDC team develops methods and tools aimed at understanding and controlling the dynamic behavior of systems in their environment (including external or internal disturbances, interaction with other cooperative or non-cooperative systems, failures, etc.).

This objective implies modeling the system and its environment, analyzing its behavior by defining criteria adapted to the desired performance, and designing various control loops required to monitor and control the system in order to meet and optimize these performance requirements. This context, which is very general to automatic control, is more particularly applied to the field of aeronautical and space systems and subsystems.

More specifically, research on modeling includes:

• development of knowledge models which are instrumental in the aeronautical or space preliminary design phases, based on general mechanics, flight mechanics, and space mechanics;

• models identification when measured data on previously designed system or subsystem are available;

• model reduction and substitution models when the physical phenomenon to be modeled does not lend itself to analytical characterization.

Research in systems analysis and synthesis of control laws includes:

• definition of system performance criteria and robustness to uncertainties, as well as design of algorithms associated with computing these criteria on increasingly complex models (analysis);

• optimization of the control law on these analysis criteria (synthesis).

These research activities lead to theoretical results, numerical methods and tools, and also the design and construction of experimental demonstrators. They are conducted within the framework of national and international programs with major industrial and academic players in the aerospace field.