Towards Certifiably Safe Nonlinear Control with Sensor and Dynamics Uncertainties

Sarah Dean

Ph.D.,
UC Berkeley

Seminar Information

Seminar Series
Dynamic Systems & Controls

Seminar Date - Time
December 3, 2021, 3:00 pm
-
4:00


Abstract

Modern nonlinear control theory seeks to endow systems with properties such as stability and safety. Despite its successful deployment in various domains, uncertainty remains a significant challenge, while data offers a potential solution. In this talk, I will discuss two settings of data-driven nonlinear control, where uncertainty arises from 1) the dynamics model and 2) the sensing model. I will introduce robust control synthesis procedures based on Control Lyapunov and Control Barrier Functions. Using this framework, we will show data-dependent guarantees. These results allow us to quantify tolerable error in learned models, and they can serve as a guide for designing sample-efficient data collection approaches. Based on joint work with Aaron Ames, Ryan Cosner, Victor Dorobantu, Ben Recht, Andrew Taylor, and Yisong Yue.

Speaker Bio

Dr. Dean just completed her PhD in EECS at UC Berkeley.

Dr. Dean study the interplay between optimization, machine learning, and dynamics in real-world systems. Hrt research focuses on understanding the fundamentals of data-driven methods for control and decision-making, inspired by applications ranging from robotics to recommendation systems. You can learn more by reading her dissertation or watching her dissertation talk.

At UC Berkeley, she was advised by Ben Recht and affiliated with BAIR and BCCI. Dr. Dean was funded by a Berkeley fellowship and the NSF graduate research fellowship.A founding member of Graduates for Engaged and Extended Scholarship in computing and Engineering (GEESE) & interned with Canopy in Boston, MA during Summer 2019.

Dr. Dean graduated from the University of Pennsylvania in 2016, where she studied electrical engineering and math and was fortunate to work with professors Daniel Lee and Daniel Koditschek. During her time at Penn, she worked as a teaching assistant for several math and engineering courses and became engaged in service learning through her involvement with the West Philadelphia Tutoring Project at Civic House.