A General Safety-Certified Cooperative Control Architecture for Interconnected Intelligent Surface Vehicles With Applications to Vessel Train

This paper considers cooperative control of interconnected intelligent surface vehicles (ISV) moving in a complex water surface containing multiple static/dynamic obstacles. Each ISV is subject to control force and moment constraints, in addition to internal model uncertainties and external disturbances induced by wind, waves and currents. A general safety-certified cooperative control architecture capable of achieving various collective behaviors such as consensus, containment, enclosing, and flocking, is proposed. Specifically, a distributed motion generator is used to generate desired reference signals for each ISV. Robust-exact-differentiators-based (RED-based) extended state observers (ESOs) are designed for recovering unknown total disturbances in finite time. With the aid of control Lyapunov functions, input-to-state safe high order control barrier functions and RED-based ESOs, constrained quadratic optimization problems are formulated to generate optimal surge force and yaw moment without violating the input, stability, safety constraints. In order to facilitate real-time implementations, a one-layer recurrent neural network is employed to solve the constrained quadratic optimization problem on board. It is proved that all tracking errors of the closed-loop system are uniformly ultimately bounded and the multi-ISV system is input-to-state safe. An example is given to substantiate the effectiveness of the proposed general safety-certified cooperative control architecture.