Traction Control Systems (TCSs) are nowadays widely adopted and they enhance both the performance and the safety of motorcycles. The use of TCSs is beneficial for both novice and experienced motorcyclists, as they allow for a safer, carefree and more accessible ride by relieving the pilot from adapting the throttle to the road and motorbike conditions, while at the same time maximizing the achieved longitudinal acceleration. The rider exposure to incidents makes the safety trait of TCSs an invaluable improvement that compensates for the rider’s inability to react quickly enough to wheel slip. Considering their effectiveness, many countries are even evaluating the mandatory installation of traction control systems on all newly registered motorcycles. Despite TCSs being well developed for road bikes, little research has been made towards their application to off-road vehicles that, instead, would benefit the most since being inherently employed on loose, low-friction surfaces. State-of-the-art traction controllers directly aim to reduce the slipping of the driven wheel, typically estimated through wheel velocity encoders. A further challenge for off-road motorcycles is that they usually consist of very simple vehicles equipped with very basic and low-cost sensors and the installation of new sensors is not always viable; both for economical reasons and sport regulations that explicitly forbid the installation of certain sensors. A wide span of users is therefore uncovered from this safety improvement. This dissertation addresses the development of a TCS for such motorbikes, paying attention to the estimation issues arising from the limited sensors perimeter. In addition, this thesis develops modules that tackle specific features of an off-road environment, consistently with the limited sensor set. Such characteristics can be problematic for the traction controller. The solution is attained through the classic steps of model-based control design, supported by the test driver’s feedback for the fine tuning of the controllers proposed. We conducted experimental sessions aimed at validating the implementation of the proposed analysis.
Oggigiorno i sistemi di controllo trazione (TCSs) sono ampiamente adottati migliorando sia la prestazione che la sicurezza delle motociclette. L’uso del controllo di trazione beneficia sia i motociclisti esperti sia i novizi, permettendo una guida più sicura, spensierata ed accessibile, sollevando il pilota dal continuo adattamento del comando dell’acceleratore alle condizioni della strada e della moto, ottenendo allo stesso tempo la massima accelerazione possibile. La maggior esposizione dei piloti ai rischi rende il controllo di trazione un miglioramento imprescindibile dato che compensa l’inabilità di una rapida reazione ai cambiamenti di slip. La sua efficacia è provata dal fatto che diversi paesi stanno valutando di rendere il controllo di trazione un sistema obbligatorio sulle nuove motociclette. Nonostante il TCS sia ben sviluppato per le motociclette da strada, ancora poca ricerca è stata fatta per la sua applicazione sulle motociclette da fuoristrada, che beneficerebbero molto di questo sistema date le superfici sdrucciolevoli dove sono utilizzate. I controllori di trazione allo stato dell’arte puntano direttamente a ridurre lo slittamento della ruota motrice, tipicamente stimato mediante encoder di velocità ruota. Un ulteriore problema delle motociclette da fuoristrada è che spesso sono veicoli molto elementari, equipaggiati con sensori semplici ed a basso costo e spesso l’installazione di nuovi sensori non è possibile per motivi economici e di regolamenti. Ad una grande quantità di utenti rimane quindi precluso l’utilizzo del controllo di trazione. Questa tesi si concentra sulla realizzazione di un TCS per questo tipo di motociclette, prestando particolare attenzione ai problemi riguardanti le stime causati dall’insieme ristretto di sensori. Inoltre, sono stati sviluppati moduli specifici per gestire situazioni, in ambito fuoristrada, problematiche per un controllo di trazione con perimetro di sensori limitato. La soluzione è raggiunta attraverso le classiche fasi della sintesi del controllore model-based, coadiuvate dai riscontri del pilota di test per la messa a punto dei controllori. Una sessione sperimentale ha permesso di validare l’implementazione delle analisi proposte.
Analysis and experimental validation of a traction control for off-road motorcycles with a limited sensor set
Martinelli, Nicolas;Ongaro, Daniele
2021/2022
Abstract
Traction Control Systems (TCSs) are nowadays widely adopted and they enhance both the performance and the safety of motorcycles. The use of TCSs is beneficial for both novice and experienced motorcyclists, as they allow for a safer, carefree and more accessible ride by relieving the pilot from adapting the throttle to the road and motorbike conditions, while at the same time maximizing the achieved longitudinal acceleration. The rider exposure to incidents makes the safety trait of TCSs an invaluable improvement that compensates for the rider’s inability to react quickly enough to wheel slip. Considering their effectiveness, many countries are even evaluating the mandatory installation of traction control systems on all newly registered motorcycles. Despite TCSs being well developed for road bikes, little research has been made towards their application to off-road vehicles that, instead, would benefit the most since being inherently employed on loose, low-friction surfaces. State-of-the-art traction controllers directly aim to reduce the slipping of the driven wheel, typically estimated through wheel velocity encoders. A further challenge for off-road motorcycles is that they usually consist of very simple vehicles equipped with very basic and low-cost sensors and the installation of new sensors is not always viable; both for economical reasons and sport regulations that explicitly forbid the installation of certain sensors. A wide span of users is therefore uncovered from this safety improvement. This dissertation addresses the development of a TCS for such motorbikes, paying attention to the estimation issues arising from the limited sensors perimeter. In addition, this thesis develops modules that tackle specific features of an off-road environment, consistently with the limited sensor set. Such characteristics can be problematic for the traction controller. The solution is attained through the classic steps of model-based control design, supported by the test driver’s feedback for the fine tuning of the controllers proposed. We conducted experimental sessions aimed at validating the implementation of the proposed analysis.File | Dimensione | Formato | |
---|---|---|---|
Tesi_Martinelli_Ongaro.pdf
non accessibile
Descrizione: Tesi
Dimensione
12.71 MB
Formato
Adobe PDF
|
12.71 MB | Adobe PDF | Visualizza/Apri |
Executive_summary_Martinelli_Ongaro.pdf
non accessibile
Descrizione: Executive summary
Dimensione
1.38 MB
Formato
Adobe PDF
|
1.38 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/10589/197401