Performance of Single Phase DVR Using Conventional PI AND ANN Controllers

Abstract: This paper proposes conventional proportional integral (PI) controller and artificial neural network (ANN) controller of single phase dynamic voltage restorer (DVR) to enhance the performance of distribution system. The main contribution of this work is designing an ANN controller by using the data obtained from the conventional PI controller for single phase DVR and comparing the performance of both the controllers. For designing the conventional PI controller, two different PI controllers are used for d-axis and q-axis respectively. The optimal PI control parameters are determined using the Zeigler-Nichols method. All the values obtained from the PI controller are then used to train the ANN. The Levenberg Marquardt back propagation algorithm is used as the ANN controller’s training algorithm. The goal of ANN controller is to minimize the error between the actual value and the reference value. After the completion of ANN training the ANN controller is installed in place of PI controller. To validate the effectiveness of the PI control and ANN control, three scenarios are included i.e. sag, swell and harmonics conditions. The ANN training and system design is done using MATLAB/Simulink.