Performance-Guaranteed Control for Euler-Lagrange Systems with Uncertain Direction Actuator Faults

This paper investigates the prescribed performance tracking control problem for uncertain Euler-Lagrange systems subjected to unknown disturbances, parameters, and actuator faults with unpredictable jumps. These factors pose significant challenges to system performance. By employing some functional transformations, the performance-guaranteed problem is reformulated as a boundary problem of new variables. Unlike conventional methods, the proposed neural network (NN)-based fault-tolerant performance-guaranted method ensures that the performance boundary of the tracking error can be flexibly adjusted through the selection of appropriate parameters in the constraint function, avoiding the need to reset the control structure and desired trajectory. To handle unknown system parameters, radial basis function neural network (RBFNN) technology is used for approximation. Additionally, the challenges posed by unknown parameters and unpredictable actuator faults jumps are mitigated through a novel recursive contradiction argument, leveraging Nussbaum functions and adaptive estimation techniques. A key contribution of this work is the assurance of both global symmetric and semi-global asymmetric performance behaviors.