A Convex Approach to Data-Driven Optimal Control via Perron–Frobenius and Koopman Operators

This article is about the data-driven computation of optimal control for a class affine deterministic nonlinear systems. We assume that dynamical system model not available, and only information dynamics available in form time-series data. provide convex formulation problem (OCP) system. The relies on duality result system’s stability theory involving density function Perron–Frobenius operator. formulate OCP as an infinite-dimensional optimization program. finite-dimensional approximation recent advances made Koopman operator’s computation, which dual to Simulation results are presented demonstrate application developed framework.