Motorcycle Riding Simulator: How to Estimate Robustly the Rider's Action?

Lamri Nehaoua (1) , Hichem Arioui (1) , Said Mammar (1)

1 IBISC - Informatique, Biologie Intégrative et Systèmes Complexes

This paper deals with a motorcycle riding simulator and addresses two key issues: 1) reconstruction of rider's action, considered as the main input to the simulation system and 2) design an appropriate force feedback, on the handlebar, emulating a tire-road contact. To answer the first challenge, a Walcott-Zak based sliding-mode observer is designed for the rider torque estimation. The reconstructed torque is used as the main control of the virtual motorcycle dynamic model, in order to actuate the simulator's platform. The steering system is modeled as a haptic display subjected to a couple of action-reaction torques: rider and tire-road dynamics. Besides, a torque feedback is implemented to compensate the lack of the real tire-road contact. The control approach is based on a robust tracking problem of a reference steering angle by using $\mathcal{H}_\infty$ optimization technique.

Mots clés: Unknown Input Estimation; Sliding Mode Observer; Riding Simulator; Unknown Input Estimation.

Domaines: Automatique / Robotique

Lamri Nehaoua, Hichem Arioui, Said Mammar. Motorcycle Riding Simulator: How to Estimate Robustly the Rider's Action?. IEEE Transactions on Vehicular Technology, 2013, pp.80-88. ⟨10.1109/TVT.2012.2215052⟩. ⟨hal-00725426⟩