Various rider models have been proposed that provide control inputs for the simulation of motorcycle dynamics. However, those models are mostly used to simulate production motorcycles, so they assume that all motions are in the linear region such as those in a constant radius turn. As such, their performance is insufficient for simulating racing motorcycles that experience quick acceleration and braking. Therefore, this study proposes a new rider model for racing simulation that incorporates Nonlinear Model Predictive Control. In developing this model, it was built on the premise that it can cope with running conditions that lose contact with the front wheels or rear wheels so-called "endo" and "wheelie", which often occur during running with large acceleration or deceleration assuming a race. For the control inputs to the vehicle, we incorporated the lateral shift of the rider's center of gravity in addition to the normally used inputs such as the steering angle, throttle position, and braking force. We compared the performance of the new model with that of the conventional model under constant radius cornering and straight braking, as well as complex braking and acceleration in a single (hairpin) corner that represented a racing run. The results showed that the new rider model outperformed the conventional model, especially in the wider range of running speed usable for a simulation. In addition, we compared the simulation results for complex braking and acceleration in a single hairpin corner produced by the new model with data from an actual race and verified that the new model was able to accurately simulate the run of actual MotoGP riders.
On multiple coverings of the infinite rectangular grid with balls of constant radius