Titel: A Journey into Optimal Control via Scientific Computing

Abstract: 

Optimal control lies at the heart of many problems in applied mathematics, where the goal is to determine policies that minimize a cost while respecting system dynamics. The two classical approaches, i.e. Pontryagin’s Maximum Principle and Bellman’s Dynamic Programming, offer complementary analytical and numerical perspectives: one grounded in necessary optimality conditions, the other in Hamilton–Jacobi–Bellman equations for the value function. In this talk, after briefly revisiting these foundational viewpoints, we will discuss how modern scientific computing can both accelerate their implementation and offer new insights. In particular, we will present examples where GPU-based (CUDA) algorithms enable large-scale experiments that make the theory applicable to real-world scenarios. Finally, we will explore some emerging connections between optimal control and optimal transport, showing how a gradient flow formulation can be used to approximate solutions to mean field games.

Online link: 

https://oeaw-ac-at.zoom.us/j/69785906386?pwd=zlAikyw4mVUhQRaqY7j8SRBPvexNo0.1

Meeting ID: 697 8590 6386
Passcode: 8qT0WX