Efficient zonotope reduction via hyperplane zonotopic contractors for uncertainty representation
Rahma Bengamra, Soheib Fergani, Carine Jauberthie · IFAC-PapersOnLine · 2025
This paper introduces an enhanced zonotope reduction method to address the trade-off between computational efficiency and precision in uncertain dynamical systems, a step forward in the management of conservatism. Zonotopes are widely used for bounded uncertainty representation due to their favorable algebraic properties, but their complexity escalates with the number of generators, requiring effective reduction strategies. Here, an improved approach based on the Hyperplan Zonotopic Contractor (HZC) reduction method is introduced, which consistently achieves higher accuracy with fewer generators. Unlike interval-based methods, this approach operates directly on zonotopes, preserving their geometric properties while optimizing computational performance. Various cases simulations demonstrate the efficiency of the proposed method, particularly for generator counts within practical ranges. The preliminary results are promising and suggest that the proposed method could serve as a foundation for future advances in precise and scalable zonotope reduction techniques.