Globally Robust Adaptive Critic Based Neuro-Integral Terminal Sliding Mode Technique with UDE for Nonlinear Systems

摘要

This paper presents an innovative neural network based optimal integral terminal sliding mode control framework for stabilization of an uncertain affine class of nonlinear systems. In literature, an uncertainty and disturbance estimator (UDE) has been successfully integrated with sliding mode control to diminish the influence of unknown system uncertainties through a low pass filter. However, this estimator causes an initial control signal to shoot up to a very large value which may undesirably affect the functioning of the connected actuators and sensing devices. To this end, this paper utilizes a single network continuous adaptive critic optimal technique which would significantly reduce the initial large control magnitude generated by UDE technique. To ensure the global robustness, an exponential bilateral decay function is employed in the proposed control law. Moreover, the excellent approximation characteristic of radial basis feed-forward neural network is also exploited to estimate the partial non-linear dynamics. The closed loop stability is also guaranteed with the proposed technique through Lyapunov's principle. Finally, two practical examples are simulated to state the efficacy of the proposed method and comparison with the prior methods is also provided in the presented work.