Neuro-Adaptive Integral Sliding-Mode Consensus Control for Multi-ASVs With Unknown Nonlinearity via Dissipative Disturbance Estimation

This article investigates a dissipative disturbance-estimator (DDE)-based neuro-adaptive integral sliding-mode control (ISMC) for multiple autonomous surface vehicles (ASVs) subject to unknown nonlinearities and mismatched external disturbances. A memory-based integral sliding manifold is first developed by integrating radial basis function neural networks (RBFNNs) and state-dependent input matrices to approximate nonlinear dynamics. A disturbance observer (DOB) is then designed to estimate mismatched disturbances and actively compensate for their effects. On this basis, an adaptive ISMC is constructed with dynamic laws that estimate the upper bounds of neural network approximation errors. Furthermore, an improved integral inequality and a novel asymmetric Lyapunov-Krasovskii functional (ALKF) are introduced, leading to dissipative performance criteria expressed as linear matrix inequalities (LMIs), from which disturbance-rejection gains are obtained. Finally, numerical simulations on multi-ASVs, together with a comparative example, validate the effectiveness and superiority of the proposed approach.