Title:  Safety Embedded Optimal Decision Making and Control via Barrier States

Committee: 

Dr. Theodorou, Advisor

Dr. Sadegh, Co-Advisor 

Dr. Coogan, Chair

Dr. Vamvoudakis

Dr. Vela

Abstract: The objective of the proposed research is to confront the trade-off between safety restrictions and performance through designing the appropriate mathematical tools used to develop provably safe and robust optimal control and planning for general safety-critical dynamical systems and path constraints. In this thesis, aiming to develop algorithms that achieve safety and optimality simultaneously, I first pursue the design of embedded barrier states (BaS) as a means of integrating safety into performance objectives. BaS are subsequently used with various optimal control frameworks, such as nonlinear quadratic regulators (NLQR), differential dynamic programming (DDP) and model predictive control (MPC) to develop novel algorithms that produce safety-aware and robust decision making. Additionally, the proposed algorithms are considered for uncertain safety critical problems in which game theoretic and stochastic formulations are utilized. This proposal suggests studying the employment of barrier states methods in different safety-critical application such as multi-agent systems, data-driven representations, and perception. Furthermore, I propose to apply the aforementioned safety-embedded methodology to Lyapunov analyses as well as constrained dynamical optimization. The proposed work and algorithms are to be deployed on real-world systems.