Compound learning tracking control of a switched fully-submerged hydrofoil craft

Abstract This paper addresses the tracking control of a fully-submerged hydrofoil craft (FSHC) on the horizontal plane. The motions of a FSHC can be classified into three cruising conditions, namely the hulborne, the taking off, and the foilborne. Considering that the FSHC has the different model dynamics and disturbances between the three cruising conditions, we treat the FSHC model as a switched system between the hullborne and the foliborne conditions, in which the taking off condition is conveniently omitted and deemed as the fast transient. The compound learning technique is introduced to offset the uncertainties, and it is comprised of two parts. First, the fuzzy logic system (FLS) is employed to approximate the unknown model dynamics in each degree of freedom. Based on a serial-parallel estimation model (SPEM), the composite adaptive laws are fabricated to estimate the weights of the FLS by combining the tracking errors with the prediction errors between the estimation model and the dynamic loop of the FSHC. Second, the disturbance observer (DOB) is fabricated to estimate the compound disturbance, which is comprised of the approximation error of the FLS and the disturbance. Based on the compound learning, we develop the switched adaptive backstepping control laws of the FSHC. By employing the average dwell-time method, all the tracking and the estimation errors in the switched system are proved to be semi-globally uniformly ultimately bounded (SGUUB). Finally, the numerical experiment is conducted to verify the effectiveness of the proposed scheme.