Timezone: »

Robust exploration in linear quadratic reinforcement learning
Jack Umenberger · Mina Ferizbegovic · Thomas Schön · Håkan Hjalmarsson

Tue Dec 10 05:10 PM -- 05:15 PM (PST) @ West Ballroom A + B

Learning to make decisions in an uncertain and dynamic environment is a task of fundamental performance in a number of domains. This paper concerns the problem of learning control policies for an unknown linear dynamical system so as to minimize a quadratic cost function. We present a method, based on convex optimization, that accomplishes this task ‘robustly’, i.e., the worst-case cost, accounting for system uncertainty given the observed data, is minimized. The method balances exploitation and exploration, exciting the system in such a way so as to reduce uncertainty in the model parameters to which the worst-case cost is most sensitive. Numerical simulations and application to a hardware-in-the-loop servo-mechanism are used to demonstrate the approach, with appreciable performance and robustness gains over alternative methods observed in both.

Author Information

Jack Umenberger (Uppsala University)
Mina Ferizbegovic (KTH Royal Institute of Technology)
Thomas Schön (Uppsala University)
Håkan Hjalmarsson (KTH)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors