A virtual rider for two-wheeled vehicles

Abstract

This work presents a virtual rider for the control of nonlinear motorcycle models, where the target motion is defined in terms of road path and speed profiles. The virtual rider inputs are the steering torque, the rear wheel driving/braking torque and front wheel braking torque. The virtual rider capability is assessed by guiding a nonlinear motorcycle model in very demanding cornering maneuvers, with roll angle up to 50 • , and strong accelerations which lead the vehicle to the stoppie/wheeling extreme conditions.

Gyroscopic stabilisers for powered two-wheeled vehicles

“…Other studies have confirmed this tendency, and a comprehen-sive critical discussion of the literature may be found in [17]. Accordingly, several authors have modelled rider control as a steer torque only [18][19][20][21][22][23] while some have considered the lean control too [20,24,25]. Recently, thanks to the availability of commercial multibody software, more complex rider models have been proposed.…”

A Virtual Rider for Motorcycles: Maneuver Regulation of a Multi-Body Vehicle Model

“…The resulting curve is not a system trajectory, in contrast to that provided by the dynamic inversion procedure developed in [18]. Note also that the control law in [17] is obtained using the linearization about this non-trajectory. This approach appears to work well for a class of smooth maneuvers for which that curvature does not vary too rapidly.…”

“…In a neighborhood of the desired output maneuver , we can parametrize points using (17) as depicted in Fig. 3.…”

“…A model-based predictive control is also employed in [17]. Arc length and lateral deviation from the desired path provide a notion of tracking error, much as what we propose in this work.…”