Adaptive sliding mode control of switched linear systems using disturbance observer based on the RBF neural network

This study deals with analyze of an adaptive sliding model controller for a class of switched linear systems in the context of model reference adaptive control (MRAC) using RBF neural network (RBFNN) with the aid of disturbance observer (DO). For this purpose, adaptive laws and switching rules are designed. These are constructed based on tracking error and sliding mode control, together with using time-dependent switching conceptualizations. A DO is used to estimate the external disturbance with an adaptive RBFNN which is applied to obtain the external disturbance upper bound estimation, combined with an adaptive sliding mode control (ASMC) under the identic Lyapunov stability framework. The switching rules are based on dwell time (DT) and average dwell time (ADT) switching. The ASMC updates the system dynamics so that it assures the proposed closed-loop switched linear system stability via fast switching, resulting in the form of globally uniformly ultimately bounded (GUUB) stability. The convergence of the process of updating the weights in the adaptive RBFNN and the boundedness of updated estimates of weights are satisfied. Achieving the state tracking, robustness, reducing the chattering problem and anti-disturbance performance are the main objectives. Moreover, switching rules based on the mode-dependent approaches have been developed, which can allow faster switching as compared to switching rules based on the DT and ADT. Finally, to evaluate the efficiency of the obtained theoretical results, the controller and the proposed method have been tested on the electro-hydraulic system (EHS).