PIDT: Physics-Informed Digital Twin for Optical Fiber Parameter Estimation

2026-01-12

AI Summary

This work addresses the challenges of low accuracy, slow convergence, and high computational complexity in optical fiber parameter estimation by proposing a digital twin modeling approach that integrates a parameterized stepwise method with physics-informed loss functions. For the first time, physical priors are embedded directly into the digital twin framework to construct a lightweight yet highly efficient parameter estimation model. Compared to existing neural operator methods, the proposed approach achieves significantly higher estimation accuracy and faster convergence while substantially reducing model complexity, thereby demonstrating the critical role of physics-informed guidance in enhancing the performance of data-driven modeling.

Technology Category

• Search and Optimization: Sampling/Simulation-based Search
• Intelligent Robots: State Estimation
• Computer Vision: Learning & Optimization for CV

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments

Abstract

We propose physics-informed digital twin (PIDT): a fiber parameter estimation approach that combines a parameterized split-step method with a physics-informed loss. PIDT improves accuracy and convergence speed with lower complexity compared to previous neural operators.

Problem

• fiber parameter estimation
• physics-informed
• digital twin
• optical fiber
• parameter estimation

Innovation

• Physics-Informed Digital Twin
• Fiber Parameter Estimation
• Split-Step Method
• Neural Operators
• Physics-Informed Loss