Observer based active vibration control of flexible space structures with prescribed performance

This paper proposes an active vibration control approach to solve the vibration suppression problem of flexible space structures in the presence of external disturbances and parameter uncertainties. Based on the independent modal space control method, each mode can be decoupled and controlled separately. To meet the case that the velocity sensors are not available on the flexible structures, a neural network based state observer is designed for the estimation of the modal velocity. The neural network approximation is introduced into the observer design which can effectively deal with the disturbances and uncertainties in order to obtain more accurate observation results. Further, the controller and the adaptive law are derived by using the backstepping technique and prescribed performance control method with the stability of the whole closed-loop system analyzed via Lyapunov theory. Simulation results illustrate the effectiveness and vibration control performance of the proposed method.