Learning Provable Neural Network Observer for Uncertain Dynamical Systems
Abstract
In many safety-critical applications, control of uncertain dynamical systems relies on observers that estimate states and external disturbances. Neural network observers can improve estimation accuracy, but certifying their Lyapunov stability via Linear Matrix Inequality (LMI) constraints leads to large-scale semidefinite programs (SDPs) that are difficult to solve for large networks. To overcome this scalability bottleneck, we propose a novel two-stage training framework for provably stable neural network observers. Our approach decouples the optimization into a point-guided Lyapunov pre-training phase, which rapidly achieves high estimation accuracy and local stability over sampled states, followed by an LMI fine-tuning phase that efficiently satisfies a strict global Lyapunov stability certificate. We provide formal theoretical guarantees for local stability radii and probabilistic coverage over a prescribed compact error-state domain under specified regularity and sampling assumptions. Experiments on nonlinear control benchmarks and X-29 aircraft ablations show that our LMI-certified neural network observers train significantly faster than direct LMI-based methods and generalize robustly across diverse systems, achieving improved tracking accuracy over a range of observer baselines. The code is available at https://github.com/Berry-Myon/LearningNeuralNetworkObserver.
Related
- Provable Guarantees and Efficient Learning of Structural Equation Models with Latent Confounders
- Beyond Overparameterization: Provable Learning of Input-Convex Multi-Layer Polynomial Networks with Active Queries
- Network of limit sets in dynamical systems
- Seasonal Statistics of Shannon Rate in a Dynamical Poisson-Voronoi Cellular Network
- Symmetric Hermite quadrature-based balanced truncation for learning linear dynamical systems from derivative data
BibTeX
@misc{wang2026learning,
title = {{Learning Provable Neural Network Observer for Uncertain Dynamical Systems}},
author = {Zhangyi Wang and Jiaxu Liu and Song Chen and Chao Xu and Shengze Cai},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2609.30819},
eprint = {2609.30819},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.30819},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Learning Provable Neural Network Observer for Uncertain Dynamical Systems
AU - Zhangyi Wang
AU - Jiaxu Liu
AU - Song Chen
AU - Chao Xu
AU - Shengze Cai
PY - 2026
DA - 2026/10/14
PB - arXiv
DO - 10.48550/arxiv.2609.30819
UR - https://arxiv.org/abs/2609.30819
AB - In many safety-critical applications, control of uncertain dynamical systems relies on observers that estimate states and external disturbances. Neural network observers can improve estimation accuracy, but certifying their Lyapunov stability via Linear Matrix Inequality (LMI) constraints leads to large-scale semidefinite programs (SDPs) that are difficult to solve for large networks. To overcome this scalability bottleneck, we propose a novel two-stage training framework for provably stable neural network observers. Our approach decouples the optimization into a point-guided Lyapunov pre-training phase, which rapidly achieves high estimation accuracy and local stability over sampled states, followed by an LMI fine-tuning phase that efficiently satisfies a strict global Lyapunov stability certificate. We provide formal theoretical guarantees for local stability radii and probabilistic coverage over a prescribed compact error-state domain under specified regularity and sampling assumptions. Experiments on nonlinear control benchmarks and X-29 aircraft ablations show that our LMI-certified neural network observers train significantly faster than direct LMI-based methods and generalize robustly across diverse systems, achieving improved tracking accuracy over a range of observer baselines. The code is available at https://github.com/Berry-Myon/LearningNeuralNetworkObserver.
ER -