Electrified vehicles are spreading at considerable rate, as a consequence of supportive national policies, infrastructure expansion and technical advances in the areas of energy-storage systems and vehicle controls. Doubtless, electrified vehicles with multiple drives create the conditions for more sophisticated dynamic controls, which project vehicle handling, stability, safety and energy-efficiency to a higher level. Absolutely, the development of dynamic control algorithms strongly rely on vehicle modelling and simulation. Therefore, a thorough analysis about the fundamental modelling techniques, spanning from multibody to mass-spring soft-body formulation, is carried out. A generic vehicle is then replicated through the distinct modelling techniques and the implementations are compared thanks to three significant manoeuvres. The approach allows to assess accuracy, real-time readiness and suitable contexts of application of each formulation. Subsequently, a multi-actuated fully electric vehicle equipped with four in-wheel-motors is investigated from an energy-efficiency viewpoint. Three systems of actuation, namely, torque vectoring, anti-roll moment distribution via active suspensions and rear-wheel steering, in addition to a longitudinal motor torque distribution strategy, are explored through a systematic simulation campaign by using an experimentally-validated high-fidelity nonlinear vehicle model. Specifically, power-saving capabilities of single and combined actuations are assessed along a set of ramp-steer manoeuvres performed at different speeds, adherence levels and workload conditions. The analysis reveals good energy-saving capabilities related to torque vectoring and rear-wheel steering at low-to-medium lateral accelerations, and a substantial power-saving authority connected to anti-roll moment distribution at medium-to-high lateral accelerations. Also, in the presence of additional powertrain workload, longitudinal motor torque distribution can increase considerably the energy-efficiency. Afterwards, nonlinear model predictive control (NMPC) is applied to the torque vectoring and front-to-total anti-roll moment distribution control of a four-wheel-drive electric vehicle with in-wheel-motors and active suspension actuators. The NMPC cost function formulation strives to minimise the power losses due to longitudinal and lateral tyre slips and the electric powertrains, while enhancing the vehicle cornering response in steady-state and transient conditions. The capabilities of the proposed controller are evaluated through simulations along ramp steer and double-step steer manoeuvres, with and without the active anti-roll moment distribution together with torque vectoring. The results show the considerable enhancement of energy saving and vehicle stabilisation performance brought by the integration of the active suspension contribution and torque vectoring.
I veicoli elettrificati si stanno diffondendo ad un ritmo notevole, risultato di politiche nazionali favorevoli, infrastrutture in espansione e progressi tecnici nell'ambito dei sistemi di accumulo dell'energia e dei controlli di veicolo. Senza dubbio i mezzi elettrici con azionamenti multipli si prestano a controlli dinamici più sofisticati, che proiettano maneggevolezza, stabilità, sicurezza ed efficienza energetica del veicolo ad un livello superiore. Indubbiamente, lo sviluppo di algoritmi di controllo dinamici per autovetture si basa largamente su modellazione e simulazione. Perciò, si effettua una approfondita analisi delle tecniche di modellazione fondamentali, spaziando dal multibody fino alla formulazione massa-molla atta a descrivere corpi deformabili. Un veicolo generico viene quindi replicato mediante le diverse tecniche di modellazione e le implementazioni sono messe a confronto attraverso tre manovre significative. Ciò permette di valutare accuratezza, predisposizione ad impieghi in tempo reale e contesti di applicazione più adatti di ciascuna formulazione. Successivamente, un veicolo multi-attuato e con quattro motori elettrici nelle ruote è studiato dal punto di vista dell'efficienza energetica. Tre sistemi di attuazione, ossia torque vectoring, distribuzione del momento di antirollio tramite sospensioni attive e sterzo posteriore, oltre ad una strategia di ripartizione longitudinale della coppia dei motori elettrici, sono esaminate attraverso simulazioni usando un modello di veicolo ad alta fedeltà e sperimentalmente validato. Nello specifico, le potenzialità di risparmio energetico delle attuazioni singole e combinate sono valutate nel corso di una serie di manovre a rampa di sterzo eseguite per diverse condizioni di velocità, aderenza e carico dei motori. L'analisi rivela buone capacità di risparmio energetico del torque vectoring e dello sterzo posteriore a basse e medie accelerazioni laterali e notevoli potenzialità di risparmio di potenza della distribuzione del momento di antirollio alle accelerazioni laterali medio-alte. Inoltre, in caso di carico addizionale ai motori, la ripartizione longitudinale della coppia può accrescere notevolmente l'efficienza. In seguito, il controllo predittivo basato su modello non lineare è applicato al torque vectoring e al controllo di distribuzione del momento di antirollio di un veicolo elettrico con quattro motori indipendenti e sospensioni attive. La funzione di costo è volta a minimizzare le perdite di potenza e, al contempo, migliorare la risposta del veicolo sia in condizioni stazionarie, sia transitorie. Le potenzialità della strategia di controllo sono valutate su due manovre, rampa di sterzo e doppio colpo di sterzo, con e senza la distribuzione attiva del momento di antirollio a supporto del torque vectoring. I risultati mostrano notevoli miglioramenti apportati dall'integrazione del contributo delle sospensioni attive e del torque vectoring in termini di risparmio energetico e stabilizzazione del veicolo.
Energy-aware dynamic control for multi-actuated electric vehicles
2021
Abstract
Electrified vehicles are spreading at considerable rate, as a consequence of supportive national policies, infrastructure expansion and technical advances in the areas of energy-storage systems and vehicle controls. Doubtless, electrified vehicles with multiple drives create the conditions for more sophisticated dynamic controls, which project vehicle handling, stability, safety and energy-efficiency to a higher level. Absolutely, the development of dynamic control algorithms strongly rely on vehicle modelling and simulation. Therefore, a thorough analysis about the fundamental modelling techniques, spanning from multibody to mass-spring soft-body formulation, is carried out. A generic vehicle is then replicated through the distinct modelling techniques and the implementations are compared thanks to three significant manoeuvres. The approach allows to assess accuracy, real-time readiness and suitable contexts of application of each formulation. Subsequently, a multi-actuated fully electric vehicle equipped with four in-wheel-motors is investigated from an energy-efficiency viewpoint. Three systems of actuation, namely, torque vectoring, anti-roll moment distribution via active suspensions and rear-wheel steering, in addition to a longitudinal motor torque distribution strategy, are explored through a systematic simulation campaign by using an experimentally-validated high-fidelity nonlinear vehicle model. Specifically, power-saving capabilities of single and combined actuations are assessed along a set of ramp-steer manoeuvres performed at different speeds, adherence levels and workload conditions. The analysis reveals good energy-saving capabilities related to torque vectoring and rear-wheel steering at low-to-medium lateral accelerations, and a substantial power-saving authority connected to anti-roll moment distribution at medium-to-high lateral accelerations. Also, in the presence of additional powertrain workload, longitudinal motor torque distribution can increase considerably the energy-efficiency. Afterwards, nonlinear model predictive control (NMPC) is applied to the torque vectoring and front-to-total anti-roll moment distribution control of a four-wheel-drive electric vehicle with in-wheel-motors and active suspension actuators. The NMPC cost function formulation strives to minimise the power losses due to longitudinal and lateral tyre slips and the electric powertrains, while enhancing the vehicle cornering response in steady-state and transient conditions. The capabilities of the proposed controller are evaluated through simulations along ramp steer and double-step steer manoeuvres, with and without the active anti-roll moment distribution together with torque vectoring. The results show the considerable enhancement of energy saving and vehicle stabilisation performance brought by the integration of the active suspension contribution and torque vectoring.File | Dimensione | Formato | |
---|---|---|---|
report_matteo_dalboni.pdf
accesso solo da BNCF e BNCR
Tipologia:
Altro materiale allegato
Dimensione
5.5 kB
Formato
Adobe PDF
|
5.5 kB | Adobe PDF | |
phd_thesis.pdf
accesso solo da BNCF e BNCR
Tipologia:
Altro materiale allegato
Dimensione
15.11 MB
Formato
Adobe PDF
|
15.11 MB | Adobe PDF |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.14242/149614
URN:NBN:IT:UNIPR-149614