Fractional neural sliding mode control for the electro-hydraulic servo system

For the existing of complex nonlinear dynamic characteristics during controlling a large-caliber and non-equilibrium barrel´s balance and position, This paper presents a fractional neural network sliding mode control (FNSMC) strategies used in a electro-hydraulic servo system to improve the control quality of the system under the influence of non-linear aspects. For this control strategy, we use a new type of switching function called combined switching function (CSF) which consists of constant velocity switching function (CVSF) and variable speed switching function (VSSF) and adjust its switching gain based on RBF neural network on-line dynamically to obtain optimal dynamic performance. By numerical simulation, we analysis the effects of the dynamic and static characteristics for the systems based on FNSMC strategy with different switching functions. The results show that: the system based on FNSMC strategy with CSF has better robustness, effectively inhibits the chattering, improves control accuracy and reduces the control response time.