Lithium-ion batteries (LIBs) have become integral to modern energy solutions, powering electric vehicles and serving as the backbone of large-scale energy storage systems. In these applications, battery management systems (BMS) play a crucial role in ensuring safety, performance, and longevity. To enhance the functionality of BMS, advanced prognostics and health management (PHM) techniques are increasingly being adopted. These techniques leverage a variety of real-time data to monitor and predict battery degradation, thus supporting proactive maintenance and early fault detection. Despite recent advances, many current approaches remain computationally intensive and cumbersome for practical implementation. To address these challenges, this dissertation proposes an innovative PHM framework based on model-based algorithms for battery fault detection and state estimation. First, a high-fidelity simulation model is developed to capture comprehensive battery behavior, which is essential for extracting key diagnostic features and validating the new algorithms. Second, an advanced algorithm incorporating an extended Kalman filter (EKF) is introduced to facilitate real-time tracking of internal short circuit (ISC) evolution and other critical battery states. By integrating electro-thermal dynamic responses from simplified battery models with multiple measurement inputs, the proposed framework establishes a robust ISC detection mechanism that enables early identification and classification with minimal computational load. Furthermore, the algorithm enhances the joint estimation of core temperature, state-of-charge (SOC), and state-of-health (SOH) under extreme operating conditions by addressing interdependencies arising from electro-thermal aging. This integrated approach not only serves as a foundation for ISC prognosis and residual useful life (RUL) prediction in LIBs under spontaneous ISC conditions when combined with data-driven methods, but also provides both theoretical and practical advances for thermal runaway prevention. Ultimately, the framework aims to extend battery lifespan and enhance the reliability of energy storage systems in real-world applications.
Le batterie agli ioni di litio (LIB) sono diventate parte integrante delle soluzioni energetiche moderne, alimentando veicoli elettrici e costituendo la spina dorsale dei sistemi di accumulo energetico su larga scala. In queste applicazioni, i sistemi di gestione delle batterie (BMS) svolgono un ruolo cruciale nel garantire la sicurezza, le prestazioni e la durata operativa. Per migliorare la funzionalità dei BMS, si adottano sempre più tecniche avanzate di prognostica e gestione della salute (PHM). Queste tecniche sfruttano una varietà di dati in tempo reale per monitorare e prevedere il degrado della batteria, supportando così la manutenzione proattiva e il rilevamento precoce dei guasti. Nonostante i recenti progressi, molti approcci attuali rimangono complessi dal punto di vista computazionale e difficili da implementare nella pratica. Per affrontare queste sfide, questa tesi propone un innovativo framework PHM basato su algoritmi modellistici per la rilevazione dei guasti e la stima degli stati interni della batteria. Innanzitutto, viene sviluppato un modello di simulazione ad alta fedeltà per rappresentare accuratamente il comportamento della batteria, fondamentale per estrarre indicatori diagnostici chiave e validare i nuovi algoritmi. Successivamente, viene introdotto un algoritmo avanzato che integra il filtro di Kalman esteso (EKF), per consentire il tracciamento in tempo reale dell’evoluzione dei cortocircuiti interni (ISC) e di altri stati critici della batteria. Integrando le risposte dinamiche elettro-termiche provenienti da modelli semplificati con ingressi multipli di misura, il framework proposto stabilisce un meccanismo robusto per il rilevamento precoce degli ISC, permettendone l’identificazione e la classificazione con un carico computazionale minimo. Inoltre, l’algoritmo migliora la stima congiunta della temperatura interna, dello stato di carica (SOC) e dello stato di salute (SOH) anche in condizioni operative estreme, affrontando le interdipendenze derivanti dall’invecchiamento elettro-termico. Questo approccio integrato rappresenta non solo una base per la prognostica degli ISC e la previsione della vita residua utile (RUL) in presenza di ISC spontanei, in combinazione con metodi data-driven, ma costituisce anche un avanzamento teorico e pratico nella prevenzione della runaway termica. In definitiva, il framework mira ad estendere la vita utile delle batterie e ad aumentare l’affidabilità dei sistemi di accumulo energetico nelle applicazioni reali.
Advanced prognostics and health management of lithium-ion batteries using electrothermal coupling and model-based algorithms
JIA, YIQI
2024/2025
Abstract
Lithium-ion batteries (LIBs) have become integral to modern energy solutions, powering electric vehicles and serving as the backbone of large-scale energy storage systems. In these applications, battery management systems (BMS) play a crucial role in ensuring safety, performance, and longevity. To enhance the functionality of BMS, advanced prognostics and health management (PHM) techniques are increasingly being adopted. These techniques leverage a variety of real-time data to monitor and predict battery degradation, thus supporting proactive maintenance and early fault detection. Despite recent advances, many current approaches remain computationally intensive and cumbersome for practical implementation. To address these challenges, this dissertation proposes an innovative PHM framework based on model-based algorithms for battery fault detection and state estimation. First, a high-fidelity simulation model is developed to capture comprehensive battery behavior, which is essential for extracting key diagnostic features and validating the new algorithms. Second, an advanced algorithm incorporating an extended Kalman filter (EKF) is introduced to facilitate real-time tracking of internal short circuit (ISC) evolution and other critical battery states. By integrating electro-thermal dynamic responses from simplified battery models with multiple measurement inputs, the proposed framework establishes a robust ISC detection mechanism that enables early identification and classification with minimal computational load. Furthermore, the algorithm enhances the joint estimation of core temperature, state-of-charge (SOC), and state-of-health (SOH) under extreme operating conditions by addressing interdependencies arising from electro-thermal aging. This integrated approach not only serves as a foundation for ISC prognosis and residual useful life (RUL) prediction in LIBs under spontaneous ISC conditions when combined with data-driven methods, but also provides both theoretical and practical advances for thermal runaway prevention. Ultimately, the framework aims to extend battery lifespan and enhance the reliability of energy storage systems in real-world applications.| File | Dimensione | Formato | |
|---|---|---|---|
|
2025_6_Jia_Yiqi.pdf
accessibile in internet per tutti a partire dal 07/06/2028
Dimensione
19.96 MB
Formato
Adobe PDF
|
19.96 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/239117