Optimal linear quadratic control theory is applied to longitudinal and lateral control of a high-performance motorcycle. Central to the story is the use of sufficient preview of the road to obtain the full benefit available from it. The focus is on effective control near to the cornering limits of the machine, and gain scheduling according to speed and lateral acceleration is employed to ensure that the linear controller used at any time is the most appropriate to the running conditions. The motorcycle model employed and the control theory background are described briefly. Selected optimal controls and closed-loop system frequency responses are illustrated. Path-tracking simulations are discussed and results are shown. Excellent machine control near to the feasible cornering limit is demonstrated. Further work is needed to provide similarly excellent control under limit braking. (Author/publisher)
Abstract