Recursively Feasible Data-Driven Distributionally Robust Model Predictive Control With Additive Disturbances

In this letter we propose a data-driven distributionally robust Model Predictive Control framework for constrained stochastic systems with unbounded additive disturbances. Recursive feasibility is ensured by optimizing over linearly interpolated initial state constraint in combination simplified affine disturbance feedback policy. We consider moment-based ambiguity set radius the second moment of disturbance, where derive minimum number samples order to ensure user-given confidence bounds on chance constraints and closed-loop performance. This closes numerical example, highlighting performance gain satisfaction based different sample sizes.