Dynamic global proportional integral derivative sliding mode control using radial basis function neural compensator for three-phase active power filter

Abstract

An adaptive dynamic special global sliding mode controller that is based on proportional integral derivative (PID) sliding surface using radial basis function (RBF) neural network (NN) for a three- phase active power filter (APF) was presented in this paper. To overcome the problems associated with the schemes of the conventional sliding mode control, a global PID sliding manifold is introduced to realize the whole process of robustness and inhibition of the steady state error, accelerating the system response meanwhile. In addition, the nested dynamic sliding mode controller can reduce the influence of chattering that may lead to malfunction of the insulated gate bipolar transistor (IGBT) caused by sign function in the control law, achieving a better property. Moreover, owing to the parameter uncertainties and the external disturbances, a RBF neural estimator is added to eliminate the chattering phenomenon that further optimizes the performance of the system. Eventually, simulation studies in the MATLAB/SimPower Systems Toolbox verify the outstanding performance of the designed RBFNN dynamic global PID sliding mode controller in three different conditions, and some comparisons are made at the same time to demonstrate the excellent properties of the raised control method.