New physics-informed learning method for nonlinear system observers

Researchers have developed a novel physics-informed learning approach for creating Kazantzis-Kravaris (KKL) observers for nonlinear systems. This method uses a physics-informed neural network to learn the forward mapping and a standard feedforward neural network for the inverse mapping. The paper provides theoretical guarantees for the robustness of state estimation against approximation errors and system uncertainties, including non-asymptotic learning guarantees. Numerical simulations on benchmark examples demonstrate that this approach offers better generalization capabilities outside the training domain compared to existing methods. AI

IMPACT This research could lead to more robust and generalizable state estimation techniques for complex nonlinear systems.

RANK_REASON The cluster contains an academic paper detailing a new methodology for observer synthesis in nonlinear systems. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• feedforward neural network
• Kazantzis-Kravaris
• M. Umar B. Niazi
• nonlinear systems
• physics-informed neural network

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →