Adaptive neural network vibration control for a flexible vehicle-based arm with input-state constraints

In order to solve the position tracking control, vibration restraint, and complex input-state constraints problems in the actual operation process of a type of flexible vehicle-based arm system with nonlinear input constraint, time-variant state constraint, uncertain parameters, and uncertain perturbations, an effective boundary control strategy based on backstepping technique, adaptive neural network, and observer technique is put forward in this paper. First, a novel continuous, smooth, and monotonic function is to describe actuator saturation nonlinearity, which can avoid chattering efficiently. Second, potential security issue caused by limited system space is considered and described as a non-linear time-variant state constraint issue, and barrier Lyapunov function (BLF) is used to make certain that time-variant constraint is satisfied. Applying backstepping technology, an adaptive neural network control strategy is constructed to deal with the system parameter uncertainty and saturation nonlinear errors, and perturbation observers are employed to deal with perturbations uncertainty. Finally, system stabilization based on Lyapunov’s method and simulations based on Matlab show the validity of constructed control strategy.