RBFNN-Based Parameter Adaptive Sliding Mode Control for an Uncertain TQUAV With Time-Varying Mass

Abstract

In this article, a parameter adaptive sliding mode control strategy, which is based on the radial basis function neural network (RBFNN), is proposed for the trajectory tracking of an uncertain tilting quadrotor unmanned aerial vehicle (TQUAV) with time-varying mass. In this strategy, the complex uncertainties and external disturbances are considered and lumped as total disturbance terms in each channel, which can be more conveniently estimated by utilizing the RBFNN for the feedforward compensation during the controller design. Moreover, the adaptive adjustment mechanism of sliding mode manifold parameters is further explored, in which their adaptive laws can avoid monotonically increased gains. To deal with the inherent approximation errors derived from the RBFNN and the concerned time-varying mass, the parameter adaptive control method is employed, such that the impact on the evolution of the closed-loop system can be eliminated. Finally, the superior performance of the proposed control strategy can be sufficiently validated by the Lyapunov stability theory and comparative simulation results.