Observer-based control for consensus tracking of non-linear synchronous generators system using sliding mode method and a radial basis function neural network

This paper presents a novel neuro-sliding mode observer-based control strategy for addressing disturbances, model uncertainties, and unmodeled dynamics in practical multi-agent systems (MAS). The focus is on achieving consensus tracking in non-linear MAS, specifically in the context of synchronous generators. A distributed protocol based on sliding mode approach is proposed to handle unknown model structures and parameters of follower agents influenced by the dynamics of synchronous generators. To achieve consensus tracking under these conditions, a hybrid radial basis function (RBF) neural network is employed to identify the unmodeled dynamics of the follower agents. The neural network’s update law algorithm is adjusted using the errors from both the observer and the controller. The stability of the proposed method is guaranteed by employing Lyapunov theory, ensuring that the consensus error and the error between the states of the consensus error dynamic and its estimator asymptotically converge to a neighborhood of zero. To validate the theoretical results, Matlab simulations are conducted to assess the effectiveness of the proposed approach, providing evidence of its capability and practical applicability.