Neural network-controlled unified power quality conditioner for system harmonics compensation

This paper presents the design of artificial neural network (ANN)-controlled unified power quality conditioner (UPQC) for harmonics compensation in a low-voltage AC system. Simple time-domain extraction technique is used to determine the harmonic components present in the load current and source voltage. The topology is based on two 3-phase pulse width modulated (PWM) voltage source inverters (VSIs). The control of these inverters is carried out using Levenberg-Marquardt backpropagation (LMBP) ANN techniques. The simulation results show that the ANN-controlled UPQC can significantly reduce harmonic distortion in the supply system.