Interpretable Riemannian Manifold Learning AI Scheme for Application Scene Generalization with Distributed Acoustic Sensing System

Le résumé fourni par la source

Artificial intelligence (AI) is pivotal in intrusion event detection within distributed fiber optic sensing systems, which are crucial for urban pipeline monitoring and railway safety. However, current AI-based systems face high false positive/negative rates and limited generalizability due to variations in dataset quality, processing methods, and model architectures. This paper introduces an explainable deep-learning approach based on Riemannian manifolds. By exploiting the manifold space, our method identifies stable latent representations of dynamic data distributions from similar events, embedding intrusion detection within this latent manifold. The Riemannian-Transformer-OTDR (RTO) model accurately identifies identical event types across diverse scenarios, reducing misclassification rates and enhancing model stability. Field tests using a distributed fiber optic sensing system validated the effectiveness of RTO: it achieved 98.43% accuracy in a subway construction scenario and maintained 94.21% accuracy in gas pipeline monitoring, with 80.79% accuracy in a 100% noisy environment. Additionally, signal structure analysis of misclassified samples revealed insights into the model’s learning process. Our analysis of performance degradation across scenarios and noise levels further clarifies the RTO mechanisms and interpretability. Ultimately, the RTO enables precise, retraining-free intrusion event classification across multiple scenarios, offering a practical solution for rapid event detection deployment in various domains.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.