Command filtered sliding mode trajectory tracking control for unmanned airships based on RBFNN approximation

Abstract

This paper presents two sliding mode controllers to address the trajectory tracking problem of unmanned airships in the presence of unknown wind disturbance. The sliding mode controller proposed first is designed by a fast power rate reaching law(FPRRL). The disturbance is compensated by a radial basis function neural network (RBFNN). To avoid the aggressive adaptation, the controller is augmented by a command filter. The controller provides good robustness and tracking performance with no chattering under the hypothesis of ideal wind field. However, serious chattering occurs when simulation is performed under discontinuous wind field. To simulate the wind in practice, the wind field employed in the simulation is generated by the combination of a constant field and white noise. The controller is improved subsequently with an extended model to suppress the chattering induced by the white noise. The enhanced controller manipulates the derivation of system input, thus attenuating the chattering. Stability analysis shows that both controllers drive the tracking error into a controllable small region near zero. Simulations are provided to validate the performance of the proposed controllers under different wind hypothesis.