Research
Game theoretic Control
Our society is based on complex, interconnected and dynamic infrastructures, such as power, transportation and communication systems, with network structure and interacting subsystems controlled by autonomous devices and human users, generically called "agents". These systems possess the features of "complex" systems of systems (C-SoS), such as rationality and autonomy of the agents, formation of emerging behaviors, and require intelligent coordination and control actions for their safe and efficient operation.
Cooperative optimization has attracted an extraordinary amount of research attention as a methodology to let agents coordinate their actions towards a common objective, but it is ineffective for systems with self-interested agents, virtually all modern C-SoS.
A paradigm shift is necessary to ensure safe and efficient operation of C-SoS with self-interested agents. With this aim, we develop "dynamic monotone game theory" as a general mathematical framework for adaptive decision making and control in uncertain, dynamic and competitive environments that is applicable to C-SoS.
We adopt operator-theoretic tools, and integrate methods within and across dynamic game theory, multi-agent systems and distrbuted control, learning and optimization.
Applications
Smart power grids
Due to the energy transition towards distributed generation based on renewable energy sources, the operation of power systems is undergoing several paradigm shifts: decentralized device-level control, over-distributed coordination of energy sources, real-time system-level optimization, open markets involving demand response and energy storage. Applications include decentralized control of power converters in low-inertia power systems, real-time voltage control in distribution grids, optimal and distributed frequency control in transmission grids, and market-based coordination of energy supply and demand. Our research devises distributed control and optimization algorithms for smart, cyber-enabled, power grids.
Automated driving
Automated and cooperative driving hold the promise to tremendously improve safety, efficiency and profitability of road traffic in the sense of achieving near-zero traffic accidents and fatalities, preventing traffic congestions, save time and money for drivers and transportation companies, and mitigate pollution and environmental concerns. Whether it will be a success or a failure will heavily depend on technological developments of sensors, vehicle intelligence and control systems, and, most importantly, on their integration into an (artificially) intelligent automated-driving control system. Conceiving such an integration is our ultimate research goal.
Support
Research supported by the European Research Council (ERC) under project "Game theoretic Control of Complex Systems of Systems" (COSMOS), project n. 802348, 2019--2023.
