Optimal Control Seminar

Optimal Control

Optimal control is a powerful framework for making decisions over time in dynamic systems, where the aim is to find a control policy that optimizes a given objective function subject to system dynamics and constraints. It has deep roots in both engineering and economics, and provides the foundation for modern approaches in robotics, finance, and reinforcement learning. At its core, optimal control blends mathematical rigor with computational techniques to produce policies that balance immediate and future rewards while respecting the underlying system dynamics.
In this seminar, we will study optimal control through the lens of Optimal Control: Linear Quadratic Methods, a classic text that develops the linear-quadratic framework in detail. Alongside discussing key results from the book, we will also examine recent literature that extends these methods to more complex and modern settings. To complement the theory, we will implement and code up several of the algorithms covered, allowing us to bridge the gap between abstract concepts and practical applications.