Learning in Autonomous Systems

An autonomous system’s ability to learn and execute decisions in an uncertain environment involves careful consideration of its dynamic model. Gathering relevant information, processing it, and updating the model under a dynamically changing environment may cause the agent to deviate from its nominal behavior. Our research is aimed at investigating fundamental aspects in learning and control for an autonomous system, with an emphasis on learning in complex large-scale interconnected systems.

Control systems: Optimization constrained by system dynamics.

Challenges:

  • Massive nonlinearity (interconnected realistic vehicle dynamics)
  • Unknown dynamics (climate, turbulence, brain)
  • High-dimensional systems/Millions or billions of degrees of freedom (stock market, epidemics, finance)
  • Limited measurements (ocean sampling)

Opportunities:

  • Explosion of data
  • Cheap memory units
  • Experience with "traditional" control techniques