An Information-Theoretic Method for Dynamic System Identification With Output-Only Damping Estimation
arXiv:2603.29956v1 Announce Type: cross Abstract: The system identification capabilities of a novel information-theoretic method are examined here. Specifically, this work uses information-theoretic metrics and vibration-based measurements to enhance damping estimation accuracy in mechanical systems. The method refers to a key limitation in system identification, signal processing, monitoring, and alert systems. These systems integrate various components, including sensors, data acquisition devices, and alert mechanisms. They are designed to operate in an environment to calculate key parameters such as peak accelerations and duration of high acceleration values. The current operational modal identification methods, though, suffer from limitations related to obtaining poor damping estimates
View PDF
Abstract:The system identification capabilities of a novel information-theoretic method are examined here. Specifically, this work uses information-theoretic metrics and vibration-based measurements to enhance damping estimation accuracy in mechanical systems. The method refers to a key limitation in system identification, signal processing, monitoring, and alert systems. These systems integrate various components, including sensors, data acquisition devices, and alert mechanisms. They are designed to operate in an environment to calculate key parameters such as peak accelerations and duration of high acceleration values. The current operational modal identification methods, though, suffer from limitations related to obtaining poor damping estimates due to their empirical nature. This has a significant impact on alert warning systems. This occurs when their duration is misestimated; specifically, when using the vibration amplitudes as an indicator of danger alerts for monitoring systems in damage or anomaly detection scenarios. To this end, approaches based on the Shannon entropy and the Kullback-Leibler divergence concept are proposed. The primary objective is to monitor the vibration levels in near real-time and provide immediate alerts when predefined thresholds are exceeded. In considering the proposed approach, both new real-world data from the multi-axis simulation table at the University of Bath, as well as the benchmark International Association for Structural Control-American Society of Civil Engineers (IASC-ASCE) structural health monitoring problem are considered. Importantly, the approach is shown to select the optimal model, which accurately captures the correct alert duration, providing a powerful tool for system identification and monitoring.
Comments: 18 pages, 16 figures, 4 tables. Published in Journal of Dynamic Systems, Measurement, and Control (ASME), 2026. Licensed under CC BY 4.0
Subjects:
Signal Processing (eess.SP); Audio and Speech Processing (eess.AS); Systems and Control (eess.SY)
MSC classes: 93B30, 62F30, 37M10, 94A12, 68T05, 62H30, 62Cxx, 93Cxx
ACM classes: G.1.2; I.2.6; I.2.8; I.5.1; I.5.4; I.4.1; H.1.1
Cite as: arXiv:2603.29956 [eess.SP]
(or arXiv:2603.29956v1 [eess.SP] for this version)
https://doi.org/10.48550/arXiv.2603.29956
arXiv-issued DOI via DataCite (pending registration)
Journal reference: Journal of Dynamic Systems, Measurement, and Control, Vol. 148, September 2026, 051009
Related DOI:
https://doi.org/10.1115/1.4071371
DOI(s) linking to related resources
Submission history
From: Marios Impraimakis [view email] [v1] Tue, 31 Mar 2026 16:23:48 UTC (5,682 KB)
Sign in to highlight and annotate this article
Conversation starters
Daily AI Digest
Get the top 5 AI stories delivered to your inbox every morning.
More about
modelbenchmarkannounce
Resolve.ai Alternative: Open Source AI for Incident Investigation
Key Takeaway: Resolve.ai is a $1B-valued AI SRE platform used by Coinbase, DoorDash, and Salesforce — but pricing requires contacting sales with no public pricing page. Aurora is an open source (Apache 2.0) alternative that delivers autonomous AI investigation with sandboxed cloud execution, infrastructure graphs, and knowledge base search — completely free and self-hosted. What is Resolve.ai? Resolve.ai is an AI-powered autonomous SRE platform founded in 2024 by Spiros Xanthos (former SVP at Splunk, co-creator of OpenTelemetry ) and Mayank Agarwal. It raised $125M in Series A at a reported $1 billion valuation , backed by Lightspeed and Greylock with angels including Fei-Fei Li and Jeff Dean. Resolve.ai positions as "machines on call for humans" — a multi-agent AI system that autonomously
Multi-Stage Continuous Delivery
Multi-Stage Continuous Delivery - Speaker Deck speakerdeck.com El problema con los pipelines tradicionales El concepto de Multi-Stage CD es sencillo: llevas código a prod en varias iteraciones y a través de diferentes ambientes — dev, staging, prod — con fases bien definidas: build, prepare, deploy, test, notify, rollback. Suena limpio. Y en papel, lo es. El problema es la realidad. Según el State of DevOps Report 2020, el 95% del tiempo se va en mantenimiento de pipelines, el 80% en tareas manuales, y el 90% en remediación también manual. Nadie escribe esas métricas en su README, pero todos las vivimos. Los retos concretos son tres y son los de siempre: la disponibilidad de ambientes (el clásico "no le muevan a dev que estoy probando algo" ), satisfacer dependencias externas correctamente
Knowledge Map
Connected Articles — Knowledge Graph
This article is connected to other articles through shared AI topics and tags.
More in Models
Multi-Stage Continuous Delivery
Multi-Stage Continuous Delivery - Speaker Deck speakerdeck.com El problema con los pipelines tradicionales El concepto de Multi-Stage CD es sencillo: llevas código a prod en varias iteraciones y a través de diferentes ambientes — dev, staging, prod — con fases bien definidas: build, prepare, deploy, test, notify, rollback. Suena limpio. Y en papel, lo es. El problema es la realidad. Según el State of DevOps Report 2020, el 95% del tiempo se va en mantenimiento de pipelines, el 80% en tareas manuales, y el 90% en remediación también manual. Nadie escribe esas métricas en su README, pero todos las vivimos. Los retos concretos son tres y son los de siempre: la disponibilidad de ambientes (el clásico "no le muevan a dev que estoy probando algo" ), satisfacer dependencias externas correctamente
I Tried Building My Own AI… Here’s What Actually Happened
A few days ago, I decided to stop just using AI and finally try building one myself. No full roadmap. No perfect plan. Just curiosity and the willingness to figure things out. What followed was a mix of confusion, frustration, small wins, and one big realization: Building AI is way harder—and way more rewarding—than it looks. 💡 Why I Started I’ve been using AI tools for a while, like most developers. But at some point, I kept wondering: How do these models actually connect? What’s happening behind the API calls? Can I build something like this myself? Instead of watching another tutorial, I decided to just start building. ⚙️ The Stack (What I Used) I kept things simple (or at least I tried to): AI Models via API (LLMs) A basic frontend interface Deployment on a cloud platform (like Vercel
Discussion
Sign in to join the discussion
No comments yet — be the first to share your thoughts!