Power quality improvement by UPQC using ANFIS-based hysteresis controller

In this paper, an adaptive neuro-fuzzy interference system (ANFIS) that is based on hysteresis controller is being proposed for achieving the power quality improvement. The innovatory ideas behind this methodology are the smoothness obtained with the fuzzy interpolation and the adaptability for complex problems using the neural network back propagation. In addition, the neural network renders increased control over the output voltage of the series active power filter (APF) and the output current of the shunt APF too. Here, the ANFIS is trained using the target control signals of both the series APF as well as the shunt APF and with the corresponding input source side and load side parameters of the system. During the testing time, the UPQC is controlled using the control signals that are attained from the ANFIS. With the utilisation of the proposed method, the voltage and the current perturbations are reduced and the system power quality is enhanced. The MATLAB/Simulink platforms are used to execute the proposed control technique and the presentation is examined using different types of source voltage fault conditions. The effectiveness of the proposed ANFIS-based controller is analysed through the comparison analysis with the conventional control techniques.