Data-Driven Reachability Analysis from Noisy Data

We consider the problem of computing reachable sets directly from noisy data without a given system model. Several reachability algorithms are presented for different types systems generating data. First, an algorithm over-approximated based on matrix zonotopes is proposed linear systems. Constrained introduced to provide less conservative at cost increased computational expenses and utilized incorporate prior knowledge about unknown Then we extend approach polynomial and, under assumption Lipschitz continuity, nonlinear Theoretical guarantees these in that they give proper over-approximate set containing true set. Multiple numerical examples real experiments show applicability algorithms, comparisons made between algorithms.