Distributed time-varying optimal formation control for uncertain networked marine surface vehicles with explicit safety constraints.

Abstract

This paper studies distributed time-varying optimal formation control of networked marine surface vehicles (NMSVs) with explicit safety constraints. A distributed constrained optimization and tracking control framework is proposed for uncertain NMSVs, comprising an optimization layer for constrained optimal reference generation and a local control layer for physical reference tracking. In the optimization layer, the formation task is formulated as a continuous time constrained optimization problem with time-varying local cost functions. A Newton-like logarithmic barrier update law is developed to generate time-varying optimal references while preserving constraint feasibility in the optimization layer under the stated assumptions. The relaxation parameter provides a designable initialization margin that enlarges the relaxed admissible domain to accommodate a prescribed initial state. In the local control layer, a robust finite-time tracking controller is designed to compensate for model uncertainties and drive the physical NMSVs to track the optimized references generated by the optimization layer. Theoretical analysis establishes constraint feasibility and convergence of the optimization dynamics, together with finite-time tracking of the physical NMSVs. Simulation results verify the effectiveness of the proposed framework.