The energy transition has become one of the most prominent topics in recent years, and sustainable transportation is among the sectors most affected by these changes. In this context, electric mobility is rapidly evolving toward greater efficiency and technological advancement. Consequently, there is a growing need for increasingly sophisticated battery management systems (BMS) capable of accurately estimating the state of health (SOH) and state of charge (SOC) of batteries in real time. Several highly effective diagnostic methods exist, such as electrochemical impedance spectroscopy (EIS), which are typically conducted offline using highly precise laboratory instrumentation and require relatively long acquisition time to ensure accuracy. These aspects limit the applicability of the method in real driving conditions, preventing online estimation of battery diagnostic parameters. This thesis investigates the implementation of online diagnostic tools for lithium-ion batteries (LIBs), focusing in particular on two chemistries: nickel manganese cobalt (NMC) and lithium iron phosphate (LFP). The selection of these chemistries is motivated by their widespread adoption in the automotive sector. The experimental protocol was conducted using dynamic power profiles derived from the worldwide harmonized light vehicles test procedure (WLTP), adapted according to the specifications of three selected vehicles. The core contribution of this work is the validation of online EIS estimation through a wavelet-based approach. This method enables the application of the wavelet trans form (WT) to WLTP power profiles to extract impedance-related information without interrupting normal battery operation. Furthermore, a dedicated investigation on LFP cells was conducted using differential voltage analysis (DVA). Two methods were evaluated for the identification of an electrochemical peak in the dV/dSOC profile derived under real driving conditions. The results obtained from both approaches were consistent, allowing the identification of the electrochemical signature of the cell and the corresponding SOC. The results confirm a strong correlation between the impedance estimated through wavelet analysis and that obtained from laboratory measurements. In conclusion, the integration of online EIS and DVA represents a reliable multimodal diagnostic framework that signifi cantly enhances the monitoring capabilities of next-generation BMS, ensuring high levels of safety and reliability for automotive battery packs.
La transizione energetica è uno dei temi di maggiore rilievo degli ultimi anni e tra i settori più influenzati da tali cambiamenti vi è il trasporto sostenibile. In questo contesto si sviluppa la mobilità elettrica, che mira a diventare sempre più efficiente e tecnologicamente avanzata. A tal proposito cresce la necessità di sistemi di gestione della batteria (BMS) sempre più sofisticati, in grado di stimare accuratamente e in tempo reale lo stato di salute (SOH) e lo stato di carica (SOC) delle batterie. Esistono diversi metodi diagnostici molto efficaci, come la spettroscopia di impedenza elettrochimica (EIS), i quali vengono condotti tipicamente offline mediante strumentazioni di laboratorio altamente precise ed inoltre richiedono tempi relativamente lunghi per essere eseguiti con accuratezza. Tali aspetti limitano l’applicabilità del metodo in condizioni di guida reali, non consentendo una stima online dei parametri diagnostici della batteria. Questa tesi investiga l’implementazione di strumenti di diagnostica online per batterie agli ioni di litio (LIBs), focalizzandosi in particolare su due chimiche: nickel manganese cobalto (NMC) e litio ferro fosfato (LFP). La scelta di tali chimiche è motivata dalla loro ampia diffusione nel settore automotive. Il protocollo sperimentale è stato condotto attraverso l’utilizzo di profili dinamici ricavati dal worldwide harmonized light vehicles test procedure (WLTP), adattati in funzione delle specifiche di tre veicoli selezionati. Il nucleo della ricerca è la validazione della stima online dell’EIS attraverso un approccio basato sulla trasformata wavelet (WT). Questo metodo consente di applicare la WT ai profili di potenza WLTP ed estrarre informazioni relative all’impedenza senza interrompere il normale funzionamento della batteria. Inoltre, è stato condotto uno studio specifico sulle celle LFP attraverso l’applicazione della differential voltage analysis (DVA). Sono stati analizzati due metodi per l’identificazione di un picco elettrochimico nella curva dV/dSOC ricavata in condizioni di guida reale. I risultati ottenuti risultano coerenti tra i due approcci, permettendo l’identificazione della firma elettrochimica della cella e del relativo SOC. I risultati confermano una forte correlazione tra l’impedenza stimata tramite analisi wavelet e quella derivata da misure di laboratorio. In conclusione, l’integrazione tra EIS online e DVA rappresenta un framework diagnostico multimodale e affidabile che migliora significativamente le capacità di monitoraggio della nuova generazione di BMS, garantendo elevati livelli di sicurezza e affidabilità per i pacchi batteria in ambito automotive.
Advanced online diagnostics for automotive lithium batteries
Geroni, Alessandra
2025/2026
Abstract
The energy transition has become one of the most prominent topics in recent years, and sustainable transportation is among the sectors most affected by these changes. In this context, electric mobility is rapidly evolving toward greater efficiency and technological advancement. Consequently, there is a growing need for increasingly sophisticated battery management systems (BMS) capable of accurately estimating the state of health (SOH) and state of charge (SOC) of batteries in real time. Several highly effective diagnostic methods exist, such as electrochemical impedance spectroscopy (EIS), which are typically conducted offline using highly precise laboratory instrumentation and require relatively long acquisition time to ensure accuracy. These aspects limit the applicability of the method in real driving conditions, preventing online estimation of battery diagnostic parameters. This thesis investigates the implementation of online diagnostic tools for lithium-ion batteries (LIBs), focusing in particular on two chemistries: nickel manganese cobalt (NMC) and lithium iron phosphate (LFP). The selection of these chemistries is motivated by their widespread adoption in the automotive sector. The experimental protocol was conducted using dynamic power profiles derived from the worldwide harmonized light vehicles test procedure (WLTP), adapted according to the specifications of three selected vehicles. The core contribution of this work is the validation of online EIS estimation through a wavelet-based approach. This method enables the application of the wavelet trans form (WT) to WLTP power profiles to extract impedance-related information without interrupting normal battery operation. Furthermore, a dedicated investigation on LFP cells was conducted using differential voltage analysis (DVA). Two methods were evaluated for the identification of an electrochemical peak in the dV/dSOC profile derived under real driving conditions. The results obtained from both approaches were consistent, allowing the identification of the electrochemical signature of the cell and the corresponding SOC. The results confirm a strong correlation between the impedance estimated through wavelet analysis and that obtained from laboratory measurements. In conclusion, the integration of online EIS and DVA represents a reliable multimodal diagnostic framework that signifi cantly enhances the monitoring capabilities of next-generation BMS, ensuring high levels of safety and reliability for automotive battery packs.| File | Dimensione | Formato | |
|---|---|---|---|
|
2026_03_Geroni_Tesi.pdf
accessibile in internet per tutti
Descrizione: testo tesi
Dimensione
7.57 MB
Formato
Adobe PDF
|
7.57 MB | Adobe PDF | Visualizza/Apri |
|
2026_03_Geroni_Executive_Summary.pdf
accessibile in internet per tutti
Descrizione: testo executive summary
Dimensione
833.85 kB
Formato
Adobe PDF
|
833.85 kB | 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/10589/252647