Getting noncooperative agents to cooperate:nudging and dynamic interventions

Due to the strong interconnection between modern engineering systems and their users, performance of these systems heavily rely on the user behavior. Therefore, uncoordinated user behavior can deteriorate the overall performance and entail undesired outcomes. To address this problem, this thesis studies the problem of designing suitable interventions that provide coordination among noncooperative agents/players. We investigate the development of suitable interventions in several setups and propose mechanisms that achieve a desired outcome. The first part of the thesis focuses on altering the aggregative behavior of noncooperative price-taking agents towards a desired stationary or temporal behavior. We address this problem by introducing a nudge framework, where a system regulator modifies the behavior of the agents by providing a price prediction signal. In the second part of the thesis, we focus on designing intervention mechanisms that steer the actions of noncooperative players in network games to the social optimum. We investigate different cases based on the knowledge of the system regulator on the game as well as constraints on the actions and interventions. The third part of the thesis deals with the problem of Nash equilibrium seeking in aggregative games. We develop a distributed algorithm where the players communicate to their neighboring players. The robustness and privacy preserving properties of the algorithm are also analyzed