In the recent years the automotive industry requests, even more, the possibility to perform laboratory tests on prototype vehicles by using real operating conditions. Driven from the automotive market, the electrification of the vehicles cars has taken a significant role. The scope of this thesis is to presents innovative techniques and methodologies applied both to electric vehicle and the testbench: the combination of the dynamic vehicle model with the entire testbench system allows to reach better performance and improve accuracy on real estimated electric vehicles performance. The experimental characterization of the electric motor and the engine is presented together with the real-life phenomena influencing the overall system. The results have been tested in two different use cases: the complete model of a Mild Hybrid Electric Vehicle and non-linear model of a testbench system. The former use case is a linear model where the dynamic and steady-state are modelled, including the influence of external disturbance and noise. With this approach is possible evaluating different control techniques and introduce real GPS sensor data to verify the precision of the model. Two control strategies have been analysed, implemented and compared. The latter use case is a non-linear testbench model that integrates the first use case extending the application environment and using an innovative approach. This approach applied to the system is the Engine-in-the-loop. The choice to use a complex system requests a strong advanced control strategy. For this reason, the Adaptive Model Predictive Control is implemented. Both use cases are verified and validated using standard driving cycles and evaluated through the emission value in terms of carbon dioxide calculated from fuel consumption. These studies have confirmed the possibility of the system to achieve good performance in cases of real application; in this way the engine can be tested in real-time with a vehicle model, or with the innovative test bench system, and inserted into the real process, achieving the scope of the research.
Negli ultimi anni l'industria automobilistica richiede, sempre di più, la possibilità di testare prototipi di veicoli in condizioni operative reali. Spinta dal mercato automobilistico, l'elettrificazione dei veicoli ha assunto un ruolo significativo. Questa tesi presenta tecniche e metodologie innovative applicate al veicolo elettrico e al banco di test. La combinazione del modello dinamico del veicolo con l'intero sistema del banco di prova consente di raggiungere elevate prestazioni e migliorarne la precisione. La caratterizzazione sperimentale del motore elettrico e del motore endotermico viene presentata insieme ai fenomeni fisici reali che influenzano il sistema complessivo. I risultati sono stati testati in due diversi casi d'uso: il modello completo di un veicolo elettrico ibrido e il modello non lineare di un sistema banco di test. Il primo caso d'uso è un modello lineare in cui sono modellati la statica e la dinamica includendo l'influenza di disturbi e rumore esterni. Con questo approccio è possibile valutare diverse tecniche di controllo e introdurre dati reali del sensore GPS per verificare la precisione del modello. Sono state analizzate, implementate e confrontate due strategie di controllo. Il secondo caso d'uso è un modello non lineare di un banco di test che integra il primo caso d'uso estendendone l'ambiente applicativo e utilizzando un approccio innovativo. Questo approccio applicato al sistema è l'Engine-in-the-loop. La scelta di utilizzare un sistema complesso richiede una forte strategia di controllo avanzata. Per questo motivo è stato implementato l’Adaptive Model Predictive Controller. Entrambi i casi d'uso vengono verificati e validati utilizzando i cicli di guida standard e valutati attraverso il valore di emissione in termini di anidride carbonica calcolata a partire dal consumo di carburante. Questi studi hanno confermato le possibilità del sistema di raggiungere buone prestazioni in casi di applicazione reale; in questo modo il motore può essere testato in real-time con su un modello del veicolo, o con l’innovativo sistema del banco di test, e inserito nel processo reale, scopo della ricerca.
Study of intelligent test bench for development and project of electric systems for traction of innovative vehicles / Fanesi, Marika. - (2021 Mar 18).
Study of intelligent test bench for development and project of electric systems for traction of innovative vehicles
FANESI, MARIKA
2021-03-18
Abstract
In the recent years the automotive industry requests, even more, the possibility to perform laboratory tests on prototype vehicles by using real operating conditions. Driven from the automotive market, the electrification of the vehicles cars has taken a significant role. The scope of this thesis is to presents innovative techniques and methodologies applied both to electric vehicle and the testbench: the combination of the dynamic vehicle model with the entire testbench system allows to reach better performance and improve accuracy on real estimated electric vehicles performance. The experimental characterization of the electric motor and the engine is presented together with the real-life phenomena influencing the overall system. The results have been tested in two different use cases: the complete model of a Mild Hybrid Electric Vehicle and non-linear model of a testbench system. The former use case is a linear model where the dynamic and steady-state are modelled, including the influence of external disturbance and noise. With this approach is possible evaluating different control techniques and introduce real GPS sensor data to verify the precision of the model. Two control strategies have been analysed, implemented and compared. The latter use case is a non-linear testbench model that integrates the first use case extending the application environment and using an innovative approach. This approach applied to the system is the Engine-in-the-loop. The choice to use a complex system requests a strong advanced control strategy. For this reason, the Adaptive Model Predictive Control is implemented. Both use cases are verified and validated using standard driving cycles and evaluated through the emission value in terms of carbon dioxide calculated from fuel consumption. These studies have confirmed the possibility of the system to achieve good performance in cases of real application; in this way the engine can be tested in real-time with a vehicle model, or with the innovative test bench system, and inserted into the real process, achieving the scope of the research.File | Dimensione | Formato | |
---|---|---|---|
Tesi_Fanesi.pdf
Open Access dal 18/09/2022
Descrizione: Tesi_Fanesi
Tipologia:
Tesi di dottorato
Licenza d'uso:
Creative commons
Dimensione
4.3 MB
Formato
Adobe PDF
|
4.3 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.