Adaptive actuator fault and the abnormal value reconstruction for marine vehicles with a class of Euler–Lagrange system

The actuator failure in a Marine Vehicle (MV) possibly happens because of the complicated marine environment. This paper proposes an Adaptive Generalized Actuator Fault Reconstruction (AGAFR) algorithm based on an observer for a class of Euler–Lagrange systems, of which the actuator fault and environmental disturbance are combined into a generalized actuator fault for reconstruction uniformly. The adaptive method that handles the nonlinear term reduces the range of fault errors by about 62% and the convergence time by over 46% compared to the algorithm already in the literature. The stabilization term is introduced into the algorithm to effectively reduce the chatter of a high-gain algorithm. The reconstructed error is uniformly convergent to an adjustable boundary using a Lyapunov analysis. Both the adaptive method’s advantage and the stabilizing effect are demonstrated through comparative simulation.