Robust Adaptive Sliding Mode Control of Nonlinear Systems Using Neural Network

Many higher order complex nonlinear systems with unknown parameters cannot be expressed in the so called linear-in-the-unknown-parameters form. This brings a huge obstacle to the application of adaptive control to estimate the unknown parameters. Instead of estimating each of the unknown parameters of a function, a feed-forward neural network (NN) can approximate the whole function due to its universal approximation property. In this study, an adaptive neural network integral fast terminal sliding mode controller (NN-IFTSMC) has been proposed for a general n-th order nonlinear systems with parametric uncertainties and external disturbances. The closed loop system has been proven to be bounded near the origin using Lyapunov criterion. The designed control scheme was applied to a quadrotor (UAV) and an excellent trajectories' tracking were obtained.