Data‐driven predictive control for a class of uncertain control‐affine systems

This article studies a data-driven predictive control for class of control-affine systems which is subject to uncertainty. With the accessibility finite sample measurements uncertain variables, we aim find controls are feasible and provide superior performance guarantees with high probability. results into formulation stochastic optimization problem (P), intractable due unknown distribution uncertainty variables. By developing distributionally robust framework, present an equivalent yet tractable reformulation (P). Further, propose efficient algorithm that provides online suboptimal solutions To illustrate effectiveness proposed approach, consider highway speed-limit problem. We then develop set speed allow us prevent traffic congestion Finally, employ resulting method on simulator this approach numerically.