Neural network based dynamic surface second order sliding mode control for AUVs

Abstract

In this paper, a novel neural network based dynamic surface second order sliding mode control algorithm is proposed for three-dimensional trajectory tracking control of autonomous underwater vehicles (AUVs) with modeling errors under external disturbances. The controller designed is capable of strengthening robustness of the system and attenuates inherent chattering of classical sliding mode control effectively. An innovative neural network compensator is designed to counteract effects of modeling errors, furthermore, the norm of the ideal weighting vector in neural network system is regarded as the estimated parameter, such that there is only one parameter needs to be adjusted. Meanwhile, the effect of external disturbances is handled by means of hyperbolic tangent function. As a result, the Lyapunov based stability analysis is provided to guarantee semi-global uniform boundedness of all closed-loop signals. Verification of the effectiveness of the proposed algorithm is done through simulation results.