Distributed MPC-Based Robust Collision Avoidance Formation Navigation of Constrained Multiple USVs

This article is devoted to tackling the robust collision avoidance formation navigation problem for a class of multiple unmanned surface vehicles (multi-USVs), where the USVs are modeled as underactuated nonlinear systems subject to state and input constraints. Furthermore, unknown external disturbances are taken into consideration, as they stem from diverse uncertainties associated with environmental loadings and are encountered in various practical situations. Considering the inherent nonlinear dynamics and the state and input constraints, a new kind of distributed model predictive control (MPC) based controllers is developed to achieve collision free formation navigation. Specifically, in light of the unavailability of accurate USV dynamics caused by unknown external disturbances, time-delay observers are constructed to estimate these disturbances. This estimation process facilitates the creation of a reliable predictive model, which in turn enables the design of an effective MPC controller. Subsequently, a class of distributed collision avoidance MPC formation navigation control strategies is presented and utilized such that the control inputs of USVs can be determined synchronously. It is shown that the time-delay observers can effectively estimate the external disturbances and lead to satisfactory performance of the distribute MPC based controller. At last, numerical experiments are conducted to validate the effectiveness of the present control strategy and to demonstrate its advantages over existing approaches.