Multimodel fuzzy predictor-based distributed game-theoretic formation control of underactuated autonomous surface vehicles

This paper investigates a distributed game-theoretic formation control problem for underactuated autonomous surface vehicles (ASVs) with potentially conflicting individual objectives. A distributed game-theoretic formation control framework is developed, in which formation coordination based on the Nash equilibrium (NE) is integrated with the control design for underactuated ASV dynamics. First, a distributed observer is designed to provide the position estimates required for the distributed implementation of the NE seeking law. Second, a multimodel fuzzy predictor-based method is developed, where two monitoring signals select suitable predictors from two banks constructed using two sets of predefined control gain values to estimate the unknown control gains and nonlinear model uncertainties in the ASV dynamics. On this basis, a distributed game-theoretic formation controller is constructed to achieve distributed formation control in the NE sense. Finally, Lyapunov analysis establishes the semiglobal uniform ultimate boundedness of all closed-loop error signals. Simulation results demonstrate the effectiveness of the proposed method in achieving the desired formation. • Noncooperative game-based ASV formation considers conflicting individual objectives. • Distributed Nash equilibrium seeking is investigated for underactuated ASVs. • Multimodel fuzzy predictors estimate unknown control gains and model uncertainties.